Targeted Delivery of the ErbB2/HER2 Tumor Antigen to Professional APCs Results in Effective Antitumor Immunity
http://www.100md.com
免疫学杂志 2005年第9期
Abstract
Activation of T cells by professional APCs that present peptide epitopes of tumor-associated Ags is critical for the induction of cell-mediated immunity against tumors. To facilitate targeted delivery of the ErbB2 (HER2, neu) tumor Ag to APCs in vivo, we have generated chimeric proteins that contain the extracellular domain of CTLA-4 for binding to B7 molecules on the APC surface, which is genetically fused to a human ErbB2 fragment as an antigenic determinant. Bacterially expressed CTLA-4-ErbB2 fusion protein and a similar molecule harboring in addition the translocation domain of Pseudomonas exotoxin A as an endosome escape function displayed specific binding to B7-expressing cells, followed by protein internalization and intracellular degradation. Vaccination of BALB/c mice with the fusion proteins resulted in the induction of ErbB2-specific CD8+ T cells and CTL-dependent protection from subsequent challenge with ErbB2-expressing but not ErbB2-negative murine renal carcinoma cells. In a therapeutic setting, injection of CTLA-4-ErbB2 protein vaccines caused rejection of established ErbB2-expressing tumors. Thereby, immunological memory was induced, leading to long-term systemic immunity and protection against rechallenge several months later. Our results demonstrate that these chimeric protein vaccines are effective tools for the induction of ErbB2-specific, T cell-mediated immunity.
Introduction
Cancer vaccination aims at the induction of specific humoral and/or cellular immune responses against Ags, which are expressed exclusively or overexpressed by the tumor. Such tumor-associated Ags can be mutated self-Ags, viral Ags associated with the tumor, so-called cancer-testis or cancer-germline Ags, differentiation Ags, or Ags overexpressed in tumors but also present in many normal tissues (1, 2). The epidermal growth factor receptor-related receptor tyrosine kinase ErbB2 (HER2, neu) belongs to this latter class of unmodified self-Ags. Gene amplification and ErbB2 overexpression have been observed in many human tumors, including breast and ovarian cancers, non-small cell lung cancers, and cancers of the head and neck, and have been linked with cancer development and progression (3, 4, 5). With humanized mAb Herceptin, an ErbB2-specific reagent is in clinical use for immunotherapy of breast cancer (6), and alternative Ab-based therapeutics are under development (7, 8).
The generation of an active immune response to ErbB2 in patients represents a promising alternative to passive immunotherapy (9). Several HLA-A2-restricted ErbB2 peptide epitopes could be defined, including epitopes recognized by ex vivo-stimulated CTLs on ovarian, breast, renal cell carcinoma, gastric cancer, and melanoma cells (10, 11). Preexistent CD4+ T cell responses to ErbB2 have been demonstrated in breast cancer patients (12), and endogenous Ab responses have been found in patients with breast, ovarian, colon, and prostate cancers (12, 13, 14, 15). This suggests that tolerance to ErbB2 could be overcome without inducing destructive autoimmunity. In patients immunized with ErbB2 peptide vaccines, long-lived Th and CTL responses could be demonstrated (16, 17). However, so far, clinical responses associated with ErbB2 peptide vaccination have not been reported.
In contrast to mixtures of defined peptides, large protein fragments in principle contain a broader number of potential Th and CTL epitopes suitable for presentation by different MHC alleles, but to be effective, the large protein fragments require proper processing and presentation by professional APCs in vivo. In experimental models, DNA vaccines encoding different fragments of ErbB2 or the rat homologue Neu protected against challenge with transplanted tumors and inhibited carcinogenesis in transgenic animals (9, 18, 19, 20). Thereby, combination with cytokines or other immunostimulatory activities enhanced the success of vaccination (21, 22). In contrast, immunization with ErbB2 or Neu protein fragments in the absence of additional adjuvants was either ineffective or resulted only in partial protection against transplantable tumors, requiring the combination with specific immunostimulatory activities or carriers to enhance effectiveness (9, 23, 24, 25).
To investigate whether specific targeting of an antigenic ErbB2 protein fragment to APCs in vivo can enhance the induction of ErbB2-specific T cell responses and increase antitumoral activity in the absence of adjuvants or synthetic carriers, in the present study, we have generated chimeric ErbB2 fusion proteins, which contain a soluble form of CTLA-4 as a cell-binding domain for specific interaction with B7 proteins expressed on APCs. Targeting of these ErbB2 protein vaccines to B7-expressing cells and specific uptake were confirmed in vitro. Induction of ErbB2-specific T cell responses in vivo and protective and therapeutic effects of ErbB2 protein vaccines were investigated in immunocompetent BALB/c mice using ErbB2-expressing murine renal carcinoma cells as a model.
Materials and Methods
Cells and culture conditions
Human Raji B cell lymphoma cells, THP-1 acute monocytic leukemia cells (American Type Culture Collection), and primary murine splenocytes were cultured in RPMI 1640 medium and supplemented with 10% FBS, 2 mM glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin. Murine D2SC/1 dendritic cells were grown in DMEM containing the same supplements (26). RPMI 1640 medium for CD80-transfected THP-1 cells in addition contained 1 mg/ml G418 (27), medium for murine renal carcinoma cells 0.2 mg/ml Zeocin (Renca-lacZ), or 0.2 mg/ml Zeocin and 0.5 mg/ml G418 (Renca-lacZ/CD80 and Renca-lacZ/ErbB2) (28).
Antibodies
Anti-CD16/CD32 mAb 2.4G2, anti-human CTLA-4 mAb BNI3, and FITC-conjugated anti-mouse IgG were purchased from BD Pharmingen, PE-conjugated anti-mouse IgG Ab was purchased from Dianova, and PE-conjugated anti-CD8 mAb 5H10 was purchased from Caltag Laboratories. Anti-Myc tag mAb 9E10, anti-human ErbB2 mAb FRP5, anti-CD4 mAb YTS191, FITC-conjugated anti-CD4 mAb GK1.5, Cy5-conjugated anti-CD8 mAb YTS169, and FITC-conjugated anti-IFN- mAb XMG1.2 were purified from hybridoma culture supernatants and conjugated following standard protocols.
Construction and bacterial expression of CTLA-4-ErbB2 fusion proteins
A cDNA fragment encoding the extracellular domain of human CTLA-4 (aa 1-125; CTLA-4125) (27) was inserted downstream of an isopropyl -D-thiogalactoside-inducible tac promoter and 5' of a synthetic sequence coding for a Myc tag and a cluster of six histidine residues (His tag) in the bacterial expression vector pSW5 (7). Based on the resulting construct, plasmids pSW5-CTLA-4125-ErbB2222 encoding the CTLA-4125 domain fused in frame to amino acids 1-222 of human ErbB2 followed by C-terminal Myc and His tags (C-ErbB2) and pSW5-CTLA-4125-ETA252-366-ErbB2222 encoding a similar protein, in addition harboring the translocation domain (aa 252-366) (29) of Pseudomonas exotoxin A (ETA)3 between CTLA-4 and ErbB2 fragments (C-E-ErbB2), were derived by stepwise assembly. Similarly, for expression of Myc- and His-tagged ErbB2222 control protein, the pSW5 derivative pSW5- ErbB2222 was constructed.
For bacterial expression of recombinant proteins, plasmids were transformed into codon-optimized Escherichia coli strain BL21-CodonPlus (DE3)-RIL (Stratagene). Single colonies were grown at 37°C to an OD550 of 0.8 to 1.0 in Luria-Bertani medium containing 100 μg/ml ampicillin. Protein expression was induced with 1 mM isopropyl -D-thiogalactoside for 2 h at 37°C before cells were harvested by centrifugation at 5000 x g for 20 min at 4°C. Cell pellets from 1 liter of culture were resuspended in 35 ml of PBS containing 8 M urea. Cells were disrupted by ultrasonification, lysates were incubated for 45 min at room temperature (RT), and clarified by centrifugation at 40,000 x g for 30 min at 4°C. Recombinant proteins were purified via binding of their polyhistidine tags to Ni2+-saturated Chelating Sepharose (Amersham Biosciences) equilibrated with PBS and 8 M urea and elution with PBS, 8 M urea, and 250 mM imidazole. Fractions containing recombinant proteins were identified by SDS-PAGE and immunoblotting with anti-Myc-tag mAb, pooled, and dialyzed against PBS and 400 mM L-arginine at 4°C. L-Arginine was then removed by stepwise dilution of dialysis buffer with PBS.
Binding assays
Binding of purified C-ErbB2, C-E-ErbB2, and ErbB2222 proteins to different cell lines expressing human or murine B7 proteins was determined by FACS analysis. A total of 5 x 105 to 1 x 106 cells was incubated with 5–6 μg of recombinant proteins for 45 min on ice, followed by incubation with 2 μg of mAb 9E10 and PE-labeled goat anti-mouse IgG for 30 min. Then cells were washed and bound proteins were detected using FACScan and FACSCalibur (BD Biosciences) flow cytometers.
Analysis of protein uptake
Raji cells were incubated with 20 μg/ml CTLA-4-ErbB2 fusion proteins for 1 h at 4°C, washed with PBS, and incubated additionally in fresh medium at 37°C. Cell aliquots were taken after 1, 2, or 4 h at 37°C and placed on poly-L-lysine-coated glass slides for 15 min. Cells were fixed for 10 min with 4% paraformaldehyde in PBS at RT, followed by permeabilization with 0.1% Triton X-100 in PBS for 5 min. Permeabilized cells were incubated for 1 h with primary ErbB2-specific Ab FRP5 in PBS containing 3% BSA, followed by incubation with FITC-coupled anti-mouse Ab. Then samples were mounted in Vectashield mounting medium (Vector Laboratories) and analyzed using a Leica TCS SL laser scanning microscope (Leica Microsystems).
Restimulation of splenocytes and evaluation of T cell responses
Female BALB/c mice of 15- to 17-g body weight (Charles River Laboratories) were vaccinated by an i.p. injection of 0.5 nmol of purified C-ErbB2 or C-E-ErbB2 proteins on days 0 and 14. Five days later, mice were sacrificed, spleens were removed, and single-cell suspensions were obtained by scraping organ tissues through a stainless steel mesh, followed by erythrocyte lysis. For detection of IFN--expressing cells by flow cytometry, 1 x 106 resuspended splenocytes/ml were stimulated for 5 h at 37°C with 10 μg/ml H-2Kd-restricted ErbB2 peptide TYLPTNASL (Thermo Electron). During the final 4 h of incubation, 10 μg/ml brefeldin A (Sigma-Aldrich) were added. Cells were washed and incubated for 10 min with 2 μg/ml anti-CD16/CD32 (Fc block) and 10 μg/ml rat serum (Sigma-Aldrich) to block unspecific binding. Then, cells were stained with Cy5-conjugated anti-CD8 mAb for 30 min at 4°C. Subsequently, cells were washed with PBS, fixed with 4% paraformaldehyde in PBS for 20 min at RT, and permeabilized with PBS, 0.1% BSA, and 0.5% saponin (Sigma-Aldrich) in the presence of 2 μg/ml anti-CD16/CD32 and 10 μg/ml rat serum. Then FITC-conjugated anti-IFN- mAb was added for 30 min at RT, cells were washed with PBS, transferred to PBS containing 1% paraformaldehyde, and analyzed using a FACSCalibur flow cytometer.
Protective and therapeutic vaccination
BALB/c mice (four to five animals per group) were vaccinated by an i.p. injection of 0.5 nmol of purified C-ErbB2, C-E-ErbB2, or ErbB2222 on days 0 and 14. On day 15, mice were inoculated with Renca-lacZ/ErbB2 or Renca-lacZ tumor cells by a s.c. injection of 5 x 106 cells into each flank. At regular intervals, two perpendicular tumor diameters were measured with a caliper and tumor volumes were calculated according to the formula: length x (width)2 x 0.4. For therapeutic vaccination, naive BALB/c mice were first inoculated with Renca-lacZ/ErbB2 cells as described above. At day 2 after tumor cell injection, when tumors were palpable, animals were treated by a s.c. injection of 0.5 nmol of purified C-ErbB2, C-E-ErbB2, or ErbB2222 into the vicinity of each tumor. Treatment was repeated on days 4 and 6, and tumor growth was followed as described above. In all cases, mice were sacrificed latest when tumor size reached 0.8 cm3.
For depletion of CD4+ or CD8+ T cell subsets, immunized mice received three i.p. injections of 100 μg of anti-CD4 mAb YTS191 or anti-CD8 mAb YTS169 on days 12, 17, and 22. Control animals were injected with 100 μg of rat IgG. Successful depletion of T cell subsets was confirmed by FACS analysis of blood samples with FITC-conjugated anti-CD4 mAb GK1.5 and PE-conjugated anti-CD8 mAb 5H10.
Long-term protection was investigated by i.v. rechallenge of surviving animals with 5 x 105 Renca-lacZ/ErbB2 tumor cells 2 mo after initial tumor cell inoculation. Cells were suspended in 100 μl of PBS and injected into the lateral tail vein. Twenty-eight days later, before mice developed apparent disease symptoms, they were sacrificed, lungs were excised, and pulmonary tumor nodules were visualized by 5-bromo-4-chloro-3-indolyl -D-galactoside (X-Gal) staining and counted as described previously (28). All animal experiments have been reviewed and approved by the appropriate government committee and were performed in accordance with the relevant guidelines and regulations.
Results
Construction and bacterial expression of chimeric CTLA-4-ErbB2 protein vaccines
The N-terminal 222-aa fragment of human ErbB2 (ErbB2222) includes the peptide epitope TYLPTNASL (residues 42-50 of mature ErbB2), which is efficiently presented by murine H-2Kd (30). For targeting ErbB2222 to professional APCs via interaction with B7 molecules on the cell surface, chimeric protein vaccines were constructed, which contain at the N terminus the extracellular domain of human CTLA-4 (aa 1-125; CTLA-4125) (Ref. 27 ; Fig. 1A). Chimeric C-ErbB2 protein consisting of CTLA-4 and ErbB2 domains, C-E-ErbB2 protein in addition harboring as an endosome escape function the translocation domain (aa 252-366) of Pseudomonas ETA (29), and recombinant ErbB2222 control protein were expressed in E. coli and purified from bacterial lysates. Identity and integrity of purified proteins were confirmed by SDS-PAGE and immunoblot analysis with CTLA-4 and ErbB2 specific Abs (Fig. 1B).
FIGURE 1. Construction and bacterial expression of chimeric CTLA-4-ErbB2 protein vaccines. A, Schematic representation of the chimeric protein C-ErbB2 consisting of aa 1-125 of human CTLA-4 fused to aa 1-222 of human ErbB2 and the similar C-E-ErbB2 protein in addition harboring the translocation domain (aa 252-366) of Pseudomonas ETA. Both fusion proteins and the ErbB2222 control protein carry C-terminal Myc (M)- and His (H)-tags. B, Immunoblot analysis of recombinant C-ErbB2 (lanes 1 and 4), C-E-ErbB2 (lanes 2 and 5), and ErbB2222 (lane 3) proteins purified from bacterial lysates. Proteins of the expected size were detected with anti-CTLA-4 (lanes 1 and 2) and FRP5 anti-ErbB2 (lanes 3–5) Abs followed by HRP-coupled secondary Ab and chemiluminescent detection. C, Binding of chimeric CTLA-4-ErbB2 proteins to B7-expressing cells. Raji and D2SC/1 cells expressing endogenous human or murine CD80 and CD86, respectively, and THP-CD80 and Renca-lacZ/CD86 cells (RLZ/CD86) transfected with human CD80 or CD86 cDNA constructs were incubated with C-ErbB2 or C-E-ErbB2 proteins as indicated (). As a control, Raji cells were also incubated with ErbB2222 protein lacking a CTLA-4 domain (upper right panel; ). Bound fusion proteins were detected by FACS analysis with Myc-tag specific Ab 9E10 and PE-conjugated secondary Ab. B7-expressing cells incubated in the absence of fusion proteins (upper panel) or B7-negative THP or Renca-lacZ cells (RLZ) incubated with fusion proteins (lower panel) were included as controls ().
Binding and cellular uptake of chimeric CTLA-4-ErbB2 proteins
Binding specificity of C-ErbB2 and C-E-ErbB2 fusion proteins was investigated by FACS analysis (Fig. 1C). Both proteins displayed strong and very similar binding to human Raji B cell lymphoma cells, which constitutively express the B7 molecules CD80 and CD86. In contrast, ErbB2222 control protein lacking the CTLA-4 domain did not bind to these cells. Specific interaction of C-E-ErbB2 protein with individual B7 molecules was confirmed using THP-CD80 and Renca-lacZ/CD86 cells transfected with human CD80 or CD86 cDNA constructs (27). Human CTLA-4 has been reported previously to interact functionally with murine B7 proteins and vice versa (31). Accordingly, binding of chimeric CTLA-4-ErbB2 proteins to murine D2SC/1 dendritic cells, which express endogenous CD80 and CD86 proteins (26), could be demonstrated for C-E-ErbB2 by FACS analysis.
The events following binding of C-ErbB2 and C-E-ErbB2 proteins to B7-expressing cells were investigated by immunofluorescence analysis with ErbB2-specific Ab FRP5 and confocal laser scanning microscopy using Raji cells as a model because these cells express CD80 and CD86 at stable levels (Fig. 2, A–I). Incubation of the cells with C-ErbB2 and C-E-ErbB2 for 1 h at 4°C resulted in binding of the proteins to the cell surface (Fig. 2, A and E). Upon shifting of the temperature to 37°C, internalization of both chimeric proteins and redistribution to intracellular compartments was observed. Thereby, disappearance of C-ErbB2 from the cell surface was rapid and already completed after 1 h (Fig. 2, B–D). Uptake of C-E-ErbB2 occurred more slowly (Fig. 2, F–H). After 4 h at 37°C, diffuse and weak intracellular staining remained, suggesting that proteolytic degradation of the proteins had occurred resulting in a loss of the ErbB2 epitope recognized by Ab FRP5. Taken together, these data demonstrate that chimeric CTLA-4-ErbB2 fusion proteins are able to specifically target B7-expressing cells. Upon binding, the molecules are internalized and degraded in intracellular compartments, a prerequisite for the generation of peptide epitopes and Ag presentation.
FIGURE 2. Chimeric CTLA-4-ErbB2 proteins are internalized upon binding to the cell surface. Raji cells were incubated with C-ErbB2 (A–D) or C-E-ErbB2 (E–H) proteins for 1 h at 4°C (A and E). Unbound proteins were removed, and cells were incubated additionally at 37°C for 1 (B and F), 2 (C and G), or 4 h (D and H). Cells were stained with ErbB2-specific Ab FRP5 followed by FITC-labeled secondary Ab and analyzed by confocal laser scanning microscopy. Control cells were incubated in the absence of fusion proteins (I). J–L, Immunohistochemical analysis of lymph node sections. C-E-ErbB2 protein was injected s.c. into the footpad of a BALB/c mouse. Two hours later, the popliteal lymph node was removed, and frozen sections were analyzed for the presence of C-E-ErbB2 protein with a Myc-tag-specific rabbit Ab followed by alkaline phosphatase (AP) coupled-secondary Ab and AP substrate (J). Staining in the absence of primary Ab is shown in K. For comparison, a similar section was incubated with Ab detecting murine MHC II (L).
To investigate targeting of chimeric CTLA-4-ErbB2 proteins to APCs in vivo, C-E-ErbB2 protein (15 μg) was injected s.c. into the footpad of a BALB/c mouse. Two hours later, the popliteal lymph node draining from the site of exposure was resected, and frozen sections were analyzed for the presence of C-E-ErbB2 protein with a Myc-tag-specific rabbit Ab. A similar staining pattern was found for C-E-ErbB2 and MHC II (Fig. 2, J–L), suggesting that the chimeric protein localized to MHC II-expressing cells such as APCs.
Protection of mice from challenge with ErbB2-expressing tumor cells
To investigate whether vaccination with chimeric CTLA-4-ErbB2 proteins can protect animals from subsequent tumor challenge, we used BALB/c-derived renal carcinoma (Renca) cells stably transfected with the E. coli -galactosidase gene and a human c-erbB2 construct as a tumor model. Despite the expression of the foreign Ags, these Renca-lacZ/ErbB2 cells form rapidly growing local tumors upon s.c. injection into immunocompetent BALB/c mice. Intravenous injection of the cells results in the formation of pulmonary tumor nodules, which can be detected upon X-Gal staining of the organs (28).
Mice were vaccinated twice by an i.p. injection of 0.5 nmol (30 μg) of C-ErbB2, C-E-ErbB2, or ErbB2222 proteins on days 0 and 14. Control animals received PBS. One day after the last vaccination, mice were challenged by a s.c. injection of tumor cells into each flank, and tumor growth was followed. A representative experiment is shown in Fig. 3A. In contrast to the control group, all animals vaccinated with C-ErbB2 remained tumor free upon challenge with Renca-lacZ/ErbB2 cells. Upon vaccination with C-E-ErbB2 protein, three of four mice rejected the tumor cells. Interestingly, the one animal of this group not able to completely reject the tumor cells developed a tumor only on one flank and at a late time point (after day 60 of the experiment). Similar results were obtained for C-ErbB2 and C-E-ErbB2 in two additional independent experiments with four animals in each treatment group (data not shown). Also, vaccination with ErbB2222 protein lacking the CTLA-4 domain enhanced antitumor immunity, resulting in protection of half of the animals tested (two of four).
FIGURE 3. A, Vaccination with CTLA-4-ErbB2 proteins protects mice from challenge with ErbB2-expressing tumor cells. Animals were immunized by an i.p. injection of C-ErbB2 (?), C-E-ErbB2 (), or ErbB2222 () on days 0 and 14 as indicated by arrows. Control animals received PBS (). On day 15, mice were inoculated with Renca-lacZ/ErbB2 tumor cells by a s.c. injection into each flank. B, Specificity of tumor rejection. Animals were immunized with C-ErbB2 or C-E-ErbB2 proteins as described above but on day 15 were challenged by s.c. injection of ErbB2-negative Renca-lacZ cells. Tumor growth was followed by caliper measurements, and mean tumor volumes were calculated.
To examine whether ErbB2-specific responses induced by the vaccines were responsible for protection, a similar experiment was performed using isogenic, ErbB2-negative but -galactosidase-expressing Renca-lacZ cells for tumor challenge. BALB/c mice were immunized as described above and were challenged by a s.c. injection of Renca-lacZ cells into each flank. Rapid tumor growth was observed at both injection sites in all animals irrespective of the kind of vaccine they had received (Fig. 3B), strongly suggesting that the observed rejection of ErbB2-expressing Renca cells was due to ErbB2-specific immune responses induced by ErbB2-containing protein vaccines.
Some of the animals that had rejected ErbB2-expressing tumor cells were rechallenged by an i.v. injection of Renca-lacZ/ErbB2 cells 2 mo after initial s.c. tumor cell inoculation to investigate whether vaccinated mice were protected against repeated tumor challenge (Fig. 4A). Whereas numerous pulmonary tumor nodules developed in naive control mice (mean number of surface tumor nodules: 168), no tumors were detectable in the lungs of the C-ErbB2-vaccinated mice. In mice initially vaccinated with C-E-ErbB2 protein, partial protection against rechallenge was observed (mean number of surface tumor nodules: 2). Also, ErbB2222-vaccinated mice that had rejected initial Renca-lacZ/ErbB2 challenge were protected against i.v. rechallenge with these cells (data not shown). In a separate experiment, mice that had rejected initial s.c. challenge with Renca-lacZ/ErbB2 cells upon protective vaccination with C-ErbB2 or C-E-ErbB2 proteins were rechallenged 4 mo after the first challenge by an i.v. injection of ErbB2-negative Renca-lacZ cells. As shown in Fig. 4B, now these animals were also protected against outgrowth of ErbB2-negative lung tumors with 50% of the mice not developing any surface lung metastases and the remaining animals displaying drastically reduced numbers of tumor nodules in comparison to naive control mice.
FIGURE 4. A, Long-term protection of vaccinated animals. Vaccinated mice that had rejected challenge with s.c. injected tumor cells were rechallenged by an i.v. injection of Renca-lacZ/ErbB2 cells 2 mo after initial tumor challenge. Twenty-eight days later, mice were sacrificed, lungs were excised, and pulmonary tumor nodules were visualized by X-Gal staining. Development of pulmonary tumor nodules in naive animals injected i.v. with Renca-lacZ/ErbB2 cells is shown as a control. B, Determinant spreading was investigated by rechallenging vaccinated mice by an i.v. injection of ErbB2-negative Renca-lacZ cells 4 mo after the initial s.c. tumor challenge. Naive animals injected with Renca-lacZ cells served as a control. Mean numbers of pulmonary tumor nodules were determined after X-Gal staining of the excised organs as described above.
These results demonstrate that vaccination with CTLA-4-ErbB2 fusion proteins can protect against repeated challenge with ErbB2-expressing tumor cells. Furthermore, vaccination and initial tumor challenge can also result in determinant spreading and subsequent immunity to an otherwise isogenic, ErbB2-negative tumor variant.
Activation of ErbB2-specific T cells
To analyze the nature of the immune responses induced by C-ErbB2 and C-E-ErbB2 protein vaccines, BALB/c mice (five animals per group) were immunized twice by an i.p. injection of 0.5 nmol of the recombinant molecules at days 0 and 14 as described above. Five days after the last vaccination, the animals were sacrificed, and splenocytes were isolated and restimulated ex vivo with ErbB2-derived peptide TYLPTNASL (30). Activated CD8+ T cells were identified by flow cytometry after double-staining of splenocytes with Abs detecting CD8 and intracellular IFN-. Representative data from one animal of each group are shown in Fig. 5A. Mean values of the absolute numbers of CD8+IFN-+ cells from all five animals of each group are shown in Fig. 5B. Splenocytes from mice vaccinated with C-ErbB2 or C-E-ErbB2 proteins contained populations of activated CD8+ T cells that displayed markedly increased IFN- production upon restimulation with the ErbB2 peptide, indicating that immunization with both chimeric CTLA-4-ErbB2 molecules induced activation of ErbB2-specific CD8+ T cells. In contrast, in vitro restimulation of splenocytes from control animals with the ErbB2 peptide did not result in an increase in CD8+IFN-+ cells.
FIGURE 5. Immunization with CTLA-4-ErbB2 proteins induces ErbB2-specific CD8+ T cells. BALB/c mice were immunized by an i.p. injection of C-ErbB2 or C-E-ErbB2 proteins on days 0 and 14. Control animals received PBS. Five days later, splenocytes were isolated and stimulated for 5 h with H-2Kd-restricted ErbB2 peptide TYLPTNASL or left untreated before flow cytometric analysis with anti-CD8 and anti-IFN- Abs. A, Increase in CD8+IFN-+ cells upon peptide stimulation of splenocytes. Representative results from one animal of each group upon restimulation in the absence (upper panels) or presence of ErbB2 peptide are shown. B, Absolute numbers of CD8+ IFN-+ splenocytes (mean values from five animals per group).
Potential induction of Ab responses by CTLA-4-ErbB2 vaccines was also investigated. Thereby only low levels of ErbB2-specific serum Abs were detected in vaccinated animals 3 wk after immunization, making it unlikely that these contribute to tumor rejection (data not shown).
To determine the contribution of individual T cell subsets to the rejection of ErbB2-expressing tumor cells in immunized animals, BALB/c mice were vaccinated with C-ErbB2 and C-E-ErbB2 proteins on days 0 and 14 as described above, followed by s.c. challenge with Renca-lacZ/ErbB2 cells into each flank on day 16. Before tumor cell inoculation, CD8+ and/or CD4+ T cell populations were depleted by injection of CD8- and/or CD4-specific Abs on day 12. Injection of Abs was repeated on days 17 and 22. Successful depletion of T cells was controlled for each animal by FACS analysis (data not shown). Tumor volumes determined at day 34 of the experiment are shown in Fig. 6. In vaccinated mice depleted of CD8+ or CD8+ and CD4+ cells, tumors grew to sizes comparable to those of the PBS control group. In contrast, no tumors developed in animals vaccinated with C-ErbB2 or C-E-ErbB2 followed by injection of control Ab or Ab-depleting CD4+ cells. These results demonstrate that CD8+ effector cells induced by protein vaccination are absolutely required for tumor rejection.
FIGURE 6. Contribution of individual T cell subsets to the rejection of ErbB2-expressing tumor cells. BALB/c mice were injected with C-ErbB2 or C-E-ErbB2 proteins or PBS as described in the legend of Fig. 5. CD4+ and/or CD8+ T cells were depleted with anti-CD8 and/or anti-CD4 Abs on days 12, 17, and 22 as indicated. Control animals received rat IgG. Success of depletion was controlled for each mouse by FACS analysis (data not shown). On day 16, mice were inoculated with Renca-lacZ/ErbB2 tumor cells by a s.c. injection, and tumor growth was followed. Mean tumor volumes at day 34 of the experiment are shown.
Therapeutic vaccination
The influence of CTLA-4-ErbB2 protein vaccines on the growth of established tumors was investigated in animals carrying Renca-lacZ/ErbB2 tumors. BALB/c mice were inoculated with tumor cells by a s.c. injection (five animals per group). On day 2, the mice were treated by a s.c. injection of 0.5 nmol of C-ErbB2, C-E-ErbB2, or ErbB2222 proteins into the vicinity of each tumor. Control animals received PBS. Treatment was repeated on days 4 and 6. For protective vaccination, an i.p. injection of the vaccines was as efficient as a s.c. injection (data not shown). Nevertheless, because of the rapid growth of Renca-lacZ/ErbB2 tumors, for therapeutic vaccination experiments we chose s.c. injection of the proteins close to the tumors to better support immune responses in the tumor vicinity. Kinetics of tumor growth until day 43 of the experiment are shown in Fig. 7A. The survival graph in Fig. 7B displays the course of the disease in the animals, which were sacrificed latest when tumor size reached 0.8 cm3. All animals in the PBS control group had to be sacrificed due to tumor size latest until day 60 of the experiment (Fig. 7). In most vaccinated mice, tumor growth continued for a few days after onset of therapy before therapeutic effects became apparent (Fig. 7A). Forty days after tumor cell inoculation, all animals that were treated with C-ErbB2 had rejected both s.c. tumors, and the mice remained tumor free until this part of the experiment was terminated on day 96 (Fig. 7B). Therapy with C-E-ErbB2 protein initially resulted in tumor rejection in two of five animals. However, in one of these mice, tumors started to regrow on day 60, and the animal had to be sacrificed on day 73 (Fig. 7B). In the remaining mice of this group, delayed tumor growth was observed. Although injection of ErbB2222 control protein resulted in delayed tumor growth in some of the animals, none of these mice was cured (Fig. 7B). Similar results were obtained in a second experiment with smaller groups, where two of three animals treated with C-ErbB2 and three of three animals treated with C-E-ErbB2 rejected both tumors, but none of the three control mice treated with PBS. When the mice cured by C-ErbB2 or C-E-ErbB2 injection were i.v. rechallenged with Renca-lacZ/ErbB2 cells 3 mo after initial immunization (day 96), the animals in contrast to naive control mice remained completely free of pulmonary metastasis (data not shown).
FIGURE 7. Treatment with CTLA-4-ErbB2 proteins results in cures of tumor-bearing animals. BALB/c mice were inoculated with Renca-lacZ/ErbB2 tumor cells by a s.c. injection into each flank. When tumors were palpable, animals were treated by a s.c. injection of purified C-ErbB2 (?), C-E-ErbB2 (), or ErbB2222 () into the vicinity of each tumor as indicated. Treatment was repeated on days 4 and 6. Control animals received PBS (). A, Kinetics of tumor growth. Tumor growth was followed by caliper measurements, and mean tumor volumes were calculated. B, Course of the disease in mice from the experiment shown in A. Tumor-bearing mice were sacrificed latest when tumor size reached 0.8 cm3 to avoid suffering of the animals.
These results demonstrate that chimeric CTLA-4-ErbB2 proteins are not only effective as protective vaccines administered to healthy animals but are also active in a therapeutic setting inducing tumor regression and long-lasting immunity in tumor-bearing animals.
Discussion
Cancer immunotherapy has evolved as a promising strategy to use Ags differentially expressed on tumor cells for directed therapy. However, although potentially more effective than mAbs, specific cancer vaccines intended to induce endogenous antitumor responses still need to be optimized for successful clinical use (32, 33). Because induction of tumor-specific T cells is critical for cell-mediated immunity against tumors, targeting of Ags to professional APCs that can deliver effective costimulatory signals to T cells might overcome some of the limitations of present vaccination approaches.
In the present study, we have investigated whether fusion proteins designed to target a 222-aa fragment of the tumor-associated ErbB2 Ag to APCs in vivo can induce ErbB2-specific, cell-mediated antitumor immunity. The ErbB2222 fragment contains a number of CTL and Th epitopes previously shown to bind to human MHC class I and class II alleles (12, 34, 35, 36). It also includes the peptide epitope TYLPTNASL, which can be presented by human HLA-A2402 (37) and by H-2Kd in BALB/c mice (30). ErbB2 has been well recognized as a target for active immunotherapy, and a number of different strategies for vaccination have been investigated (38). Several studies have demonstrated effectiveness of ErbB2 DNA vaccines in murine models (9, 18, 19, 20). However, protein-based vaccination approaches, similar to the results obtained here with unmodified ErbB2222 protein, have only rarely achieved full protection of immunized animals (9, 23, 24, 25). Therefore, to enhance effectiveness of the ErbB2222 protein vaccine, we used the extracellular domain of human CTLA-4 for specific targeting to APCs. As a natural receptor, CTLA-4 binds with high affinity to the costimulatory B7 molecules CD80 and CD86, which are expressed predominantly on APCs (39, 40). Likewise, in the present study the bacterially expressed chimeric CTLA-4-ErbB2 proteins interacted specifically with CD80 and CD86 on the cell surface, whereas no binding to the cells was observed for ErbB2222 protein lacking a CTLA-4 domain. Upon binding, the proteins were internalized by B7-expressing Raji cells used as a model and degraded in intracellular compartments, a prerequisite for the generation of peptide epitopes and Ag presentation. CTLA-4-ErbB2 protein injected into the footpad of mice localized to areas with MHC II-expressing cells in the draining lymph node, making it likely that specific targeting of the chimeric molecules to APCs was retained in vivo.
Using respective DNA vaccine constructs in murine models, previously it was demonstrated that fusion of a CTLA-4 domain to model Ags such as human IgG or influenza virus hemagglutinin does not inhibit but can enhance Ag-specific humoral and cellular immune responses (41, 42). Nevertheless, so far it remained unclear whether such CTLA-4 fusions can also induce CTLs and cure established tumors (43). In the present study, vaccination of mice with chimeric CTLA-4-ErbB2 proteins induced ErbB2-specific CD8+ T cells recognizing the immunodominant TYLPTNASL epitope. Although cytotoxicity of such cells was not formally investigated, the failure of vaccinated mice to reject the challenge of ErbB2-expressing tumor cells upon depletion of CD8+, but not upon depletion of CD4+ cells, strongly suggests that cytotoxic CD8+ T cells are the main effector population operative in the elimination of tumor cells in vivo. Because mice were vaccinated with complete fusion proteins, in vivo activation of peptide-specific T cells must have been preceded by cellular uptake of the protein vaccines and suitable processing for presentation by MHC class I.
The immune responses induced by CTLA-4-ErbB2 protein vaccines were directed specifically against ErbB2-expressing tumor cells. Vaccinated mice were protected against the initial challenge with Renca-lacZ/ErbB2 but not ErbB2-negative Renca-lacZ cells, which still express the -galactosidase Ag. This indicates that the potent antitumoral responses observed were not caused by broad, unspecific immunostimulatory activity of the vaccines because of LPS contamination of the proteins or disturbance of endogenous B7-CTLA-4 interactions. In line with this, a similar CTLA-4-NY-ESO-1 construct failed to protect mice from challenge with Renca-lacZ/ErbB2 cells but induced rejection of Renca-lacZ cells expressing the relevant NY-ESO-1 tumor Ag (unpublished data). Vaccination with CTLA-4-ErbB2 proteins and tumor rejection induced immunological memory, resulting in long-term systemic immunity and protection against subsequent rechallenge with i.v. injected Renca-lacZ/ErbB2 cells several months later. Importantly, after vaccination with CTLA-4-ErbB2 proteins and rejection of initial challenge with ErbB2-expressing cells, mice now also acquired immunity to the ErbB2-negative tumor cell variant and rejected systemic challenge of Renca-lacZ cells. Possibly, tumor cell destruction by ErbB2-specific T cells resulted in the induction or enhancement of immune responses directed against additional Ags such as the bacterial -galactosidase, which in naive animals in our BALB/c-based model failed to serve as a tumor rejection Ag. In a clinical setting, similar determinant spreading could be beneficial to achieve effective immunity also against variant tumor cells that have lost expression of the Ag used for vaccination (44).
The C-E-ErbB2 molecule, in addition to the B7-specific cell binding domain and the antigenic ErbB2 fragment, carries the translocation domain of Pseudomonas ETA as an endosome escape function (29). After binding of full-length toxin to target cells and receptor-mediated endocytosis, this domain facilitates translocation of an enzymatically active ETA fragment to the cytoplasm (45). Previously, fusion of influenza virus peptide epitopes to an enzymatically inactive ETA derivative has been used to facilitate cellular uptake and enhance MHC I presentation of the Ags (46). More recently, vaccination of mice with a DNA vaccine encoding a fusion of ETA translocation domain with human papillomavirus type 16 E7 was reported to result in a 30-fold increase in E7-specific CD8+ T cells in comparison to unmodified E7 DNA vaccine (47). In these studies, ETA fragment and Ag were not connected to an APC-targeting domain. In the present study, the C-E-ErbB2 protein and C-ErbB2 lacking the ETA domain were very similar in their efficiency to induce ErbB2-specific CD8+ T cells and protect mice from tumor challenge. This suggests that specific targeting to APCs might already be sufficient to enhance cross-presentation or delivery of target Ag to an intracellular compartment suitable for MHC I loading (48), making a separate endosome escape function dispensable.
Vaccination of BALB/c mice with DNA vaccines encoding extracellular and transmembrane domains of the rat ErbB2 homologue Neu was reported previously to protect against lethal challenge with Neu-expressing tumor cells. Thereby CD4+ T cells and granulocytes were shown to be crucial for the antitumoral response, whereas depletion of CD8+ cells only partly affected protection (19). In contrast, while most likely required during initial priming of the response (49), during the effector phase, CD4+ cells were dispensable in animals immunized with C-ErbB2 or C-E-ErbB2 proteins. However, CD8+ T cells were absolutely essential for tumor rejection.
In this study concerned with the initial characterization of chimeric CTLA-4-ErbB2 protein vaccines, we have used human ErbB2 as an Ag in a murine model. Previously, CTLs from BALB/c mice that, similar to the CD8+ T cells identified in our study, recognize the human ErbB2 epitope TYLPTNASL have been shown to also recognize and lyse cells presenting the corresponding murine ErbB2 peptide TYLPANASL (30). This indicates that cross-reactive immune responses are possible and suggests that at least for this particular epitope tolerance to endogenous murine ErbB2 might have to be overcome to mount an effective response against the human Ag. However, in this context, it is noteworthy that in our BALB/c-based model also ErbB2222 protein lacking the CTLA-4 domain induced long-term immunity upon protective vaccination, albeit in a smaller number of animals. In contrast, in a therapeutic setting, ErbB2222 treatment only resulted in delayed tumor growth in some of the mice but was unable to cure any of the animals. To further investigate the possible influence of tolerance to the chosen Ag on the efficacy of such nontargeted and APC-targeted vaccines, subsequent studies will be performed in transgenic animals such as mice transgenic for human ErbB2 (20) or oncogenically activated or normal rat Neu (50, 51). Experiments with CTLA-4-Neu constructs in BALB-neuT mice (50) have been initiated.
Taken together, our results demonstrate that bacterially expressed chimeric proteins combining a tumor Ag and specific recognition of APCs in a single molecule can induce cell-mediated, Ag-specific immune responses and tumor rejection. Although in our model recombinant ErbB2222 on its own resulted in protective immunity in some of the vaccinated mice, it was ineffective as a therapeutic reagent and could not cure tumor-bearing animals. Fusion of this protein fragment to CTLA-4 strongly enhanced its efficacy, now also resulting in rejection of established tumor grafts.
Acknowledgments
We thank Manuela St?ber for purification of mAbs and Drs. Elke J?ger and Torsten Tonn for helpful suggestions and discussions.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported in part by Wilhelm Sander-Stiftung Grant 2002.010.1.
2 Address correspondence and reprint requests to Dr. Winfried S. Wels, Chemotherapeutisches Forschungsinstitut, Georg-Speyer-Haus, Paul-Ehrlich-Stra?e 42-44, D-60596 Frankfurt am Main, Germany. E-mail address: wels@em.uni-frankfurt.de
3 Abbreviations used in this paper: ETA, exotoxin A; RT, room temperature; X-Gal, 5-bromo-4-chloro-3-indolyl -D-galactoside.
Received for publication November 19, 2004. Accepted for publication February 8, 2005.
References
Van Der Bruggen, P., Y. Zhang, P. Chaux, V. Stroobant, C. Panichelli, E. S. Schultz, J. Chapiro, B. J. Van Den Eynde, F. Brasseur, T. Boon. 2002. Tumor-specific shared antigenic peptides recognized by human T cells. Immunol. Rev. 188:51.
Scanlan, M. J., A. O. Gure, A. A. Jungbluth, L. J. Old, Y. T. Chen. 2002. Cancer/testis antigens: an expanding family of targets for cancer immunotherapy. Immunol. Rev. 188:22.
Slamon, D. J., W. Godolphin, L. A. Jones, J. A. Holt, S. G. Wong, D. E. Keith, W. J. Levin, S. G. Stuart, J. Udove, A. Ullrich, et al 1989. Studies of the HER-2/neu proto-oncogene in human breast and ovarian cancer. Science 244:707.
Klapper, L. N., M. H. Kirschbaum, M. Sela, Y. Yarden. 2000. Biochemical and clinical implications of the ErbB/HER signaling network of growth factor receptors. Adv. Cancer Res. 77:25.
Olayioye, M. A., R. M. Neve, H. A. Lane, N. E. Hynes. 2000. The ErbB signaling network: receptor heterodimerization in development and cancer. EMBO J. 19:3159.
Slamon, D. J., B. Leyland-Jones, S. Shak, H. Fuchs, V. Paton, A. Bajamonde, T. Fleming, W. Eiermann, J. Wolter, M. Pegram, J. Baselga, L. Norton. 2001. Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N. Engl. J. Med. 344:783.
Rohrbach, F., B. Gerstmayer, M. Biburger, W. Wels. 2000. Construction and characterization of bispecific costimulatory molecules containing a minimized CD86 (B7-2) domain and single chain antibody fragments for tumor targeting. Clin. Cancer Res. 6:4314.
Azemar, M., S. Djahansouzi, E. Jager, C. Solbach, M. Schmidt, A. B. Maurer, K. Mross, C. Unger, G. von Minckwitz, P. Dall, B. Groner, W. S. Wels. 2003. Regression of cutaneous tumor lesions in patients intratumorally injected with a recombinant single-chain antibody-toxin targeted to ErbB2/HER2. Breast Cancer Res. Treat. 82:155.
Foy, T. M., J. Bannink, R. A. Sutherland, P. D. McNeill, G. G. Moulton, J. Smith, M. A. Cheever, K. Grabstein. 2001. Vaccination with Her-2/neu DNA or protein subunits protects against growth of a Her-2/neu-expressing murine tumor. Vaccine 19:2598.
Fisk, B., T. L. Blevins, J. T. Wharton, C. G. Ioannides. 1995. Identification of an immunodominant peptide of HER-2/neu protooncogene recognized by ovarian tumor-specific cytotoxic T lymphocyte lines. J. Exp. Med. 181:2109.
Knutson, K. L., K. Schiffman, M. A. Cheever, M. L. Disis. 2002. Immunization of cancer patients with a HER-2/neu, HLA-A2 peptide, p369-377, results in short-lived peptide-specific immunity. Clin. Cancer Res. 8:1014.
Disis, M. L., E. Calenoff, G. McLaughlin, A. E. Murphy, W. Chen, B. Groner, M. Jeschke, N. Lydon, E. McGlynn, R. B. Livingston, et al 1994. Existent T-cell and antibody immunity to HER-2/neu protein in patients with breast cancer. Cancer Res. 54:16.
Disis, M. L., S. M. Pupa, J. R. Gralow, R. Dittadi, S. Menard, M. A. Cheever. 1997. High-titer HER-2/neu protein-specific antibody can be detected in patients with early-stage breast cancer. J. Clin. Oncol. 15:3363.
Ward, R. L., N. J. Hawkins, D. Coomber, M. L. Disis. 1999. Antibody immunity to the HER-2/neu oncogenic protein in patients with colorectal cancer. Hum. Immunol. 60:510.
McNeel, D. G., L. D. Nguyen, B. E. Storer, R. Vessella, P. H. Lange, M. L. Disis. 2000. Antibody immunity to prostate cancer associated antigens can be detected in the serum of patients with prostate cancer. J. Urol. 164:1825
Knutson, K. L., K. Schiffman, M. L. Disis. 2001. Immunization with a HER-2/neu helper peptide vaccine generates HER-2/neu CD8 T-cell immunity in cancer patients. J. Clin. Invest. 107:477.
Disis, M. L., T. A. Gooley, K. Rinn, D. Davis, M. Piepkorn, M. A. Cheever, K. L. Knutson, K. Schiffman. 2002. Generation of T-cell immunity to the HER-2/neu protein after active immunization with HER-2/neu peptide-based vaccines. J. Clin. Oncol. 20:2624.
Pupa, S. M., A. M. Invernizzi, S. Forti, E. Di Carlo, P. Musiani, P. Nanni, P. L. Lollini, R. Meazza, S. Ferrini, S. Menard. 2001.
Prevention of spontaneous neu-expressing mammary tumor development in mice transgenic for rat proto-neu by DNA vaccination. Gene Ther. 8:75.
Curcio, C., E. Di Carlo, R. Clynes, M. J. Smyth, K. Boggio, E. Quaglino, M. Spadaro, M. P. Colombo, A. Amici, P. L. Lollini, P. Musiani, G. Forni. 2003. Nonredundant roles of antibody, cytokines, and perforin in the eradication of established Her-2/neu carcinomas. J. Clin. Invest. 111:1161
Piechocki, M. P., Y. S. Ho, S. Pilon, W. Z. Wei. 2003. Human ErbB-2 (Her-2) transgenic mice: a model system for testing Her-2 based vaccines. J. Immunol. 171:5787.
Rovero, S., K. Boggio, E. D. Carlo, A. Amici, E. Quaglino, P. Porcedda, P. Musiani, G. Forni. 2001. Insertion of the DNA for the 163-171 peptide of IL-1 enables a DNA vaccine encoding p185neu to inhibit mammary carcinogenesis in Her-2/neu transgenic BALB/c mice. Gene Ther. 8:447.
Cappello, P., F. Triebel, M. Iezzi, C. Caorsi, E. Quaglino, P. L. Lollini, A. Amici, E. Di Carlo, P. Musiani, M. Giovarelli, G. Forni. 2003. LAG-3 enables DNA vaccination to persistently prevent mammary carcinogenesis in HER-2/neu transgenic BALB/c mice. Cancer Res. 63:2518(Florian Rohrbach, Robert )
Activation of T cells by professional APCs that present peptide epitopes of tumor-associated Ags is critical for the induction of cell-mediated immunity against tumors. To facilitate targeted delivery of the ErbB2 (HER2, neu) tumor Ag to APCs in vivo, we have generated chimeric proteins that contain the extracellular domain of CTLA-4 for binding to B7 molecules on the APC surface, which is genetically fused to a human ErbB2 fragment as an antigenic determinant. Bacterially expressed CTLA-4-ErbB2 fusion protein and a similar molecule harboring in addition the translocation domain of Pseudomonas exotoxin A as an endosome escape function displayed specific binding to B7-expressing cells, followed by protein internalization and intracellular degradation. Vaccination of BALB/c mice with the fusion proteins resulted in the induction of ErbB2-specific CD8+ T cells and CTL-dependent protection from subsequent challenge with ErbB2-expressing but not ErbB2-negative murine renal carcinoma cells. In a therapeutic setting, injection of CTLA-4-ErbB2 protein vaccines caused rejection of established ErbB2-expressing tumors. Thereby, immunological memory was induced, leading to long-term systemic immunity and protection against rechallenge several months later. Our results demonstrate that these chimeric protein vaccines are effective tools for the induction of ErbB2-specific, T cell-mediated immunity.
Introduction
Cancer vaccination aims at the induction of specific humoral and/or cellular immune responses against Ags, which are expressed exclusively or overexpressed by the tumor. Such tumor-associated Ags can be mutated self-Ags, viral Ags associated with the tumor, so-called cancer-testis or cancer-germline Ags, differentiation Ags, or Ags overexpressed in tumors but also present in many normal tissues (1, 2). The epidermal growth factor receptor-related receptor tyrosine kinase ErbB2 (HER2, neu) belongs to this latter class of unmodified self-Ags. Gene amplification and ErbB2 overexpression have been observed in many human tumors, including breast and ovarian cancers, non-small cell lung cancers, and cancers of the head and neck, and have been linked with cancer development and progression (3, 4, 5). With humanized mAb Herceptin, an ErbB2-specific reagent is in clinical use for immunotherapy of breast cancer (6), and alternative Ab-based therapeutics are under development (7, 8).
The generation of an active immune response to ErbB2 in patients represents a promising alternative to passive immunotherapy (9). Several HLA-A2-restricted ErbB2 peptide epitopes could be defined, including epitopes recognized by ex vivo-stimulated CTLs on ovarian, breast, renal cell carcinoma, gastric cancer, and melanoma cells (10, 11). Preexistent CD4+ T cell responses to ErbB2 have been demonstrated in breast cancer patients (12), and endogenous Ab responses have been found in patients with breast, ovarian, colon, and prostate cancers (12, 13, 14, 15). This suggests that tolerance to ErbB2 could be overcome without inducing destructive autoimmunity. In patients immunized with ErbB2 peptide vaccines, long-lived Th and CTL responses could be demonstrated (16, 17). However, so far, clinical responses associated with ErbB2 peptide vaccination have not been reported.
In contrast to mixtures of defined peptides, large protein fragments in principle contain a broader number of potential Th and CTL epitopes suitable for presentation by different MHC alleles, but to be effective, the large protein fragments require proper processing and presentation by professional APCs in vivo. In experimental models, DNA vaccines encoding different fragments of ErbB2 or the rat homologue Neu protected against challenge with transplanted tumors and inhibited carcinogenesis in transgenic animals (9, 18, 19, 20). Thereby, combination with cytokines or other immunostimulatory activities enhanced the success of vaccination (21, 22). In contrast, immunization with ErbB2 or Neu protein fragments in the absence of additional adjuvants was either ineffective or resulted only in partial protection against transplantable tumors, requiring the combination with specific immunostimulatory activities or carriers to enhance effectiveness (9, 23, 24, 25).
To investigate whether specific targeting of an antigenic ErbB2 protein fragment to APCs in vivo can enhance the induction of ErbB2-specific T cell responses and increase antitumoral activity in the absence of adjuvants or synthetic carriers, in the present study, we have generated chimeric ErbB2 fusion proteins, which contain a soluble form of CTLA-4 as a cell-binding domain for specific interaction with B7 proteins expressed on APCs. Targeting of these ErbB2 protein vaccines to B7-expressing cells and specific uptake were confirmed in vitro. Induction of ErbB2-specific T cell responses in vivo and protective and therapeutic effects of ErbB2 protein vaccines were investigated in immunocompetent BALB/c mice using ErbB2-expressing murine renal carcinoma cells as a model.
Materials and Methods
Cells and culture conditions
Human Raji B cell lymphoma cells, THP-1 acute monocytic leukemia cells (American Type Culture Collection), and primary murine splenocytes were cultured in RPMI 1640 medium and supplemented with 10% FBS, 2 mM glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin. Murine D2SC/1 dendritic cells were grown in DMEM containing the same supplements (26). RPMI 1640 medium for CD80-transfected THP-1 cells in addition contained 1 mg/ml G418 (27), medium for murine renal carcinoma cells 0.2 mg/ml Zeocin (Renca-lacZ), or 0.2 mg/ml Zeocin and 0.5 mg/ml G418 (Renca-lacZ/CD80 and Renca-lacZ/ErbB2) (28).
Antibodies
Anti-CD16/CD32 mAb 2.4G2, anti-human CTLA-4 mAb BNI3, and FITC-conjugated anti-mouse IgG were purchased from BD Pharmingen, PE-conjugated anti-mouse IgG Ab was purchased from Dianova, and PE-conjugated anti-CD8 mAb 5H10 was purchased from Caltag Laboratories. Anti-Myc tag mAb 9E10, anti-human ErbB2 mAb FRP5, anti-CD4 mAb YTS191, FITC-conjugated anti-CD4 mAb GK1.5, Cy5-conjugated anti-CD8 mAb YTS169, and FITC-conjugated anti-IFN- mAb XMG1.2 were purified from hybridoma culture supernatants and conjugated following standard protocols.
Construction and bacterial expression of CTLA-4-ErbB2 fusion proteins
A cDNA fragment encoding the extracellular domain of human CTLA-4 (aa 1-125; CTLA-4125) (27) was inserted downstream of an isopropyl -D-thiogalactoside-inducible tac promoter and 5' of a synthetic sequence coding for a Myc tag and a cluster of six histidine residues (His tag) in the bacterial expression vector pSW5 (7). Based on the resulting construct, plasmids pSW5-CTLA-4125-ErbB2222 encoding the CTLA-4125 domain fused in frame to amino acids 1-222 of human ErbB2 followed by C-terminal Myc and His tags (C-ErbB2) and pSW5-CTLA-4125-ETA252-366-ErbB2222 encoding a similar protein, in addition harboring the translocation domain (aa 252-366) (29) of Pseudomonas exotoxin A (ETA)3 between CTLA-4 and ErbB2 fragments (C-E-ErbB2), were derived by stepwise assembly. Similarly, for expression of Myc- and His-tagged ErbB2222 control protein, the pSW5 derivative pSW5- ErbB2222 was constructed.
For bacterial expression of recombinant proteins, plasmids were transformed into codon-optimized Escherichia coli strain BL21-CodonPlus (DE3)-RIL (Stratagene). Single colonies were grown at 37°C to an OD550 of 0.8 to 1.0 in Luria-Bertani medium containing 100 μg/ml ampicillin. Protein expression was induced with 1 mM isopropyl -D-thiogalactoside for 2 h at 37°C before cells were harvested by centrifugation at 5000 x g for 20 min at 4°C. Cell pellets from 1 liter of culture were resuspended in 35 ml of PBS containing 8 M urea. Cells were disrupted by ultrasonification, lysates were incubated for 45 min at room temperature (RT), and clarified by centrifugation at 40,000 x g for 30 min at 4°C. Recombinant proteins were purified via binding of their polyhistidine tags to Ni2+-saturated Chelating Sepharose (Amersham Biosciences) equilibrated with PBS and 8 M urea and elution with PBS, 8 M urea, and 250 mM imidazole. Fractions containing recombinant proteins were identified by SDS-PAGE and immunoblotting with anti-Myc-tag mAb, pooled, and dialyzed against PBS and 400 mM L-arginine at 4°C. L-Arginine was then removed by stepwise dilution of dialysis buffer with PBS.
Binding assays
Binding of purified C-ErbB2, C-E-ErbB2, and ErbB2222 proteins to different cell lines expressing human or murine B7 proteins was determined by FACS analysis. A total of 5 x 105 to 1 x 106 cells was incubated with 5–6 μg of recombinant proteins for 45 min on ice, followed by incubation with 2 μg of mAb 9E10 and PE-labeled goat anti-mouse IgG for 30 min. Then cells were washed and bound proteins were detected using FACScan and FACSCalibur (BD Biosciences) flow cytometers.
Analysis of protein uptake
Raji cells were incubated with 20 μg/ml CTLA-4-ErbB2 fusion proteins for 1 h at 4°C, washed with PBS, and incubated additionally in fresh medium at 37°C. Cell aliquots were taken after 1, 2, or 4 h at 37°C and placed on poly-L-lysine-coated glass slides for 15 min. Cells were fixed for 10 min with 4% paraformaldehyde in PBS at RT, followed by permeabilization with 0.1% Triton X-100 in PBS for 5 min. Permeabilized cells were incubated for 1 h with primary ErbB2-specific Ab FRP5 in PBS containing 3% BSA, followed by incubation with FITC-coupled anti-mouse Ab. Then samples were mounted in Vectashield mounting medium (Vector Laboratories) and analyzed using a Leica TCS SL laser scanning microscope (Leica Microsystems).
Restimulation of splenocytes and evaluation of T cell responses
Female BALB/c mice of 15- to 17-g body weight (Charles River Laboratories) were vaccinated by an i.p. injection of 0.5 nmol of purified C-ErbB2 or C-E-ErbB2 proteins on days 0 and 14. Five days later, mice were sacrificed, spleens were removed, and single-cell suspensions were obtained by scraping organ tissues through a stainless steel mesh, followed by erythrocyte lysis. For detection of IFN--expressing cells by flow cytometry, 1 x 106 resuspended splenocytes/ml were stimulated for 5 h at 37°C with 10 μg/ml H-2Kd-restricted ErbB2 peptide TYLPTNASL (Thermo Electron). During the final 4 h of incubation, 10 μg/ml brefeldin A (Sigma-Aldrich) were added. Cells were washed and incubated for 10 min with 2 μg/ml anti-CD16/CD32 (Fc block) and 10 μg/ml rat serum (Sigma-Aldrich) to block unspecific binding. Then, cells were stained with Cy5-conjugated anti-CD8 mAb for 30 min at 4°C. Subsequently, cells were washed with PBS, fixed with 4% paraformaldehyde in PBS for 20 min at RT, and permeabilized with PBS, 0.1% BSA, and 0.5% saponin (Sigma-Aldrich) in the presence of 2 μg/ml anti-CD16/CD32 and 10 μg/ml rat serum. Then FITC-conjugated anti-IFN- mAb was added for 30 min at RT, cells were washed with PBS, transferred to PBS containing 1% paraformaldehyde, and analyzed using a FACSCalibur flow cytometer.
Protective and therapeutic vaccination
BALB/c mice (four to five animals per group) were vaccinated by an i.p. injection of 0.5 nmol of purified C-ErbB2, C-E-ErbB2, or ErbB2222 on days 0 and 14. On day 15, mice were inoculated with Renca-lacZ/ErbB2 or Renca-lacZ tumor cells by a s.c. injection of 5 x 106 cells into each flank. At regular intervals, two perpendicular tumor diameters were measured with a caliper and tumor volumes were calculated according to the formula: length x (width)2 x 0.4. For therapeutic vaccination, naive BALB/c mice were first inoculated with Renca-lacZ/ErbB2 cells as described above. At day 2 after tumor cell injection, when tumors were palpable, animals were treated by a s.c. injection of 0.5 nmol of purified C-ErbB2, C-E-ErbB2, or ErbB2222 into the vicinity of each tumor. Treatment was repeated on days 4 and 6, and tumor growth was followed as described above. In all cases, mice were sacrificed latest when tumor size reached 0.8 cm3.
For depletion of CD4+ or CD8+ T cell subsets, immunized mice received three i.p. injections of 100 μg of anti-CD4 mAb YTS191 or anti-CD8 mAb YTS169 on days 12, 17, and 22. Control animals were injected with 100 μg of rat IgG. Successful depletion of T cell subsets was confirmed by FACS analysis of blood samples with FITC-conjugated anti-CD4 mAb GK1.5 and PE-conjugated anti-CD8 mAb 5H10.
Long-term protection was investigated by i.v. rechallenge of surviving animals with 5 x 105 Renca-lacZ/ErbB2 tumor cells 2 mo after initial tumor cell inoculation. Cells were suspended in 100 μl of PBS and injected into the lateral tail vein. Twenty-eight days later, before mice developed apparent disease symptoms, they were sacrificed, lungs were excised, and pulmonary tumor nodules were visualized by 5-bromo-4-chloro-3-indolyl -D-galactoside (X-Gal) staining and counted as described previously (28). All animal experiments have been reviewed and approved by the appropriate government committee and were performed in accordance with the relevant guidelines and regulations.
Results
Construction and bacterial expression of chimeric CTLA-4-ErbB2 protein vaccines
The N-terminal 222-aa fragment of human ErbB2 (ErbB2222) includes the peptide epitope TYLPTNASL (residues 42-50 of mature ErbB2), which is efficiently presented by murine H-2Kd (30). For targeting ErbB2222 to professional APCs via interaction with B7 molecules on the cell surface, chimeric protein vaccines were constructed, which contain at the N terminus the extracellular domain of human CTLA-4 (aa 1-125; CTLA-4125) (Ref. 27 ; Fig. 1A). Chimeric C-ErbB2 protein consisting of CTLA-4 and ErbB2 domains, C-E-ErbB2 protein in addition harboring as an endosome escape function the translocation domain (aa 252-366) of Pseudomonas ETA (29), and recombinant ErbB2222 control protein were expressed in E. coli and purified from bacterial lysates. Identity and integrity of purified proteins were confirmed by SDS-PAGE and immunoblot analysis with CTLA-4 and ErbB2 specific Abs (Fig. 1B).
FIGURE 1. Construction and bacterial expression of chimeric CTLA-4-ErbB2 protein vaccines. A, Schematic representation of the chimeric protein C-ErbB2 consisting of aa 1-125 of human CTLA-4 fused to aa 1-222 of human ErbB2 and the similar C-E-ErbB2 protein in addition harboring the translocation domain (aa 252-366) of Pseudomonas ETA. Both fusion proteins and the ErbB2222 control protein carry C-terminal Myc (M)- and His (H)-tags. B, Immunoblot analysis of recombinant C-ErbB2 (lanes 1 and 4), C-E-ErbB2 (lanes 2 and 5), and ErbB2222 (lane 3) proteins purified from bacterial lysates. Proteins of the expected size were detected with anti-CTLA-4 (lanes 1 and 2) and FRP5 anti-ErbB2 (lanes 3–5) Abs followed by HRP-coupled secondary Ab and chemiluminescent detection. C, Binding of chimeric CTLA-4-ErbB2 proteins to B7-expressing cells. Raji and D2SC/1 cells expressing endogenous human or murine CD80 and CD86, respectively, and THP-CD80 and Renca-lacZ/CD86 cells (RLZ/CD86) transfected with human CD80 or CD86 cDNA constructs were incubated with C-ErbB2 or C-E-ErbB2 proteins as indicated (). As a control, Raji cells were also incubated with ErbB2222 protein lacking a CTLA-4 domain (upper right panel; ). Bound fusion proteins were detected by FACS analysis with Myc-tag specific Ab 9E10 and PE-conjugated secondary Ab. B7-expressing cells incubated in the absence of fusion proteins (upper panel) or B7-negative THP or Renca-lacZ cells (RLZ) incubated with fusion proteins (lower panel) were included as controls ().
Binding and cellular uptake of chimeric CTLA-4-ErbB2 proteins
Binding specificity of C-ErbB2 and C-E-ErbB2 fusion proteins was investigated by FACS analysis (Fig. 1C). Both proteins displayed strong and very similar binding to human Raji B cell lymphoma cells, which constitutively express the B7 molecules CD80 and CD86. In contrast, ErbB2222 control protein lacking the CTLA-4 domain did not bind to these cells. Specific interaction of C-E-ErbB2 protein with individual B7 molecules was confirmed using THP-CD80 and Renca-lacZ/CD86 cells transfected with human CD80 or CD86 cDNA constructs (27). Human CTLA-4 has been reported previously to interact functionally with murine B7 proteins and vice versa (31). Accordingly, binding of chimeric CTLA-4-ErbB2 proteins to murine D2SC/1 dendritic cells, which express endogenous CD80 and CD86 proteins (26), could be demonstrated for C-E-ErbB2 by FACS analysis.
The events following binding of C-ErbB2 and C-E-ErbB2 proteins to B7-expressing cells were investigated by immunofluorescence analysis with ErbB2-specific Ab FRP5 and confocal laser scanning microscopy using Raji cells as a model because these cells express CD80 and CD86 at stable levels (Fig. 2, A–I). Incubation of the cells with C-ErbB2 and C-E-ErbB2 for 1 h at 4°C resulted in binding of the proteins to the cell surface (Fig. 2, A and E). Upon shifting of the temperature to 37°C, internalization of both chimeric proteins and redistribution to intracellular compartments was observed. Thereby, disappearance of C-ErbB2 from the cell surface was rapid and already completed after 1 h (Fig. 2, B–D). Uptake of C-E-ErbB2 occurred more slowly (Fig. 2, F–H). After 4 h at 37°C, diffuse and weak intracellular staining remained, suggesting that proteolytic degradation of the proteins had occurred resulting in a loss of the ErbB2 epitope recognized by Ab FRP5. Taken together, these data demonstrate that chimeric CTLA-4-ErbB2 fusion proteins are able to specifically target B7-expressing cells. Upon binding, the molecules are internalized and degraded in intracellular compartments, a prerequisite for the generation of peptide epitopes and Ag presentation.
FIGURE 2. Chimeric CTLA-4-ErbB2 proteins are internalized upon binding to the cell surface. Raji cells were incubated with C-ErbB2 (A–D) or C-E-ErbB2 (E–H) proteins for 1 h at 4°C (A and E). Unbound proteins were removed, and cells were incubated additionally at 37°C for 1 (B and F), 2 (C and G), or 4 h (D and H). Cells were stained with ErbB2-specific Ab FRP5 followed by FITC-labeled secondary Ab and analyzed by confocal laser scanning microscopy. Control cells were incubated in the absence of fusion proteins (I). J–L, Immunohistochemical analysis of lymph node sections. C-E-ErbB2 protein was injected s.c. into the footpad of a BALB/c mouse. Two hours later, the popliteal lymph node was removed, and frozen sections were analyzed for the presence of C-E-ErbB2 protein with a Myc-tag-specific rabbit Ab followed by alkaline phosphatase (AP) coupled-secondary Ab and AP substrate (J). Staining in the absence of primary Ab is shown in K. For comparison, a similar section was incubated with Ab detecting murine MHC II (L).
To investigate targeting of chimeric CTLA-4-ErbB2 proteins to APCs in vivo, C-E-ErbB2 protein (15 μg) was injected s.c. into the footpad of a BALB/c mouse. Two hours later, the popliteal lymph node draining from the site of exposure was resected, and frozen sections were analyzed for the presence of C-E-ErbB2 protein with a Myc-tag-specific rabbit Ab. A similar staining pattern was found for C-E-ErbB2 and MHC II (Fig. 2, J–L), suggesting that the chimeric protein localized to MHC II-expressing cells such as APCs.
Protection of mice from challenge with ErbB2-expressing tumor cells
To investigate whether vaccination with chimeric CTLA-4-ErbB2 proteins can protect animals from subsequent tumor challenge, we used BALB/c-derived renal carcinoma (Renca) cells stably transfected with the E. coli -galactosidase gene and a human c-erbB2 construct as a tumor model. Despite the expression of the foreign Ags, these Renca-lacZ/ErbB2 cells form rapidly growing local tumors upon s.c. injection into immunocompetent BALB/c mice. Intravenous injection of the cells results in the formation of pulmonary tumor nodules, which can be detected upon X-Gal staining of the organs (28).
Mice were vaccinated twice by an i.p. injection of 0.5 nmol (30 μg) of C-ErbB2, C-E-ErbB2, or ErbB2222 proteins on days 0 and 14. Control animals received PBS. One day after the last vaccination, mice were challenged by a s.c. injection of tumor cells into each flank, and tumor growth was followed. A representative experiment is shown in Fig. 3A. In contrast to the control group, all animals vaccinated with C-ErbB2 remained tumor free upon challenge with Renca-lacZ/ErbB2 cells. Upon vaccination with C-E-ErbB2 protein, three of four mice rejected the tumor cells. Interestingly, the one animal of this group not able to completely reject the tumor cells developed a tumor only on one flank and at a late time point (after day 60 of the experiment). Similar results were obtained for C-ErbB2 and C-E-ErbB2 in two additional independent experiments with four animals in each treatment group (data not shown). Also, vaccination with ErbB2222 protein lacking the CTLA-4 domain enhanced antitumor immunity, resulting in protection of half of the animals tested (two of four).
FIGURE 3. A, Vaccination with CTLA-4-ErbB2 proteins protects mice from challenge with ErbB2-expressing tumor cells. Animals were immunized by an i.p. injection of C-ErbB2 (?), C-E-ErbB2 (), or ErbB2222 () on days 0 and 14 as indicated by arrows. Control animals received PBS (). On day 15, mice were inoculated with Renca-lacZ/ErbB2 tumor cells by a s.c. injection into each flank. B, Specificity of tumor rejection. Animals were immunized with C-ErbB2 or C-E-ErbB2 proteins as described above but on day 15 were challenged by s.c. injection of ErbB2-negative Renca-lacZ cells. Tumor growth was followed by caliper measurements, and mean tumor volumes were calculated.
To examine whether ErbB2-specific responses induced by the vaccines were responsible for protection, a similar experiment was performed using isogenic, ErbB2-negative but -galactosidase-expressing Renca-lacZ cells for tumor challenge. BALB/c mice were immunized as described above and were challenged by a s.c. injection of Renca-lacZ cells into each flank. Rapid tumor growth was observed at both injection sites in all animals irrespective of the kind of vaccine they had received (Fig. 3B), strongly suggesting that the observed rejection of ErbB2-expressing Renca cells was due to ErbB2-specific immune responses induced by ErbB2-containing protein vaccines.
Some of the animals that had rejected ErbB2-expressing tumor cells were rechallenged by an i.v. injection of Renca-lacZ/ErbB2 cells 2 mo after initial s.c. tumor cell inoculation to investigate whether vaccinated mice were protected against repeated tumor challenge (Fig. 4A). Whereas numerous pulmonary tumor nodules developed in naive control mice (mean number of surface tumor nodules: 168), no tumors were detectable in the lungs of the C-ErbB2-vaccinated mice. In mice initially vaccinated with C-E-ErbB2 protein, partial protection against rechallenge was observed (mean number of surface tumor nodules: 2). Also, ErbB2222-vaccinated mice that had rejected initial Renca-lacZ/ErbB2 challenge were protected against i.v. rechallenge with these cells (data not shown). In a separate experiment, mice that had rejected initial s.c. challenge with Renca-lacZ/ErbB2 cells upon protective vaccination with C-ErbB2 or C-E-ErbB2 proteins were rechallenged 4 mo after the first challenge by an i.v. injection of ErbB2-negative Renca-lacZ cells. As shown in Fig. 4B, now these animals were also protected against outgrowth of ErbB2-negative lung tumors with 50% of the mice not developing any surface lung metastases and the remaining animals displaying drastically reduced numbers of tumor nodules in comparison to naive control mice.
FIGURE 4. A, Long-term protection of vaccinated animals. Vaccinated mice that had rejected challenge with s.c. injected tumor cells were rechallenged by an i.v. injection of Renca-lacZ/ErbB2 cells 2 mo after initial tumor challenge. Twenty-eight days later, mice were sacrificed, lungs were excised, and pulmonary tumor nodules were visualized by X-Gal staining. Development of pulmonary tumor nodules in naive animals injected i.v. with Renca-lacZ/ErbB2 cells is shown as a control. B, Determinant spreading was investigated by rechallenging vaccinated mice by an i.v. injection of ErbB2-negative Renca-lacZ cells 4 mo after the initial s.c. tumor challenge. Naive animals injected with Renca-lacZ cells served as a control. Mean numbers of pulmonary tumor nodules were determined after X-Gal staining of the excised organs as described above.
These results demonstrate that vaccination with CTLA-4-ErbB2 fusion proteins can protect against repeated challenge with ErbB2-expressing tumor cells. Furthermore, vaccination and initial tumor challenge can also result in determinant spreading and subsequent immunity to an otherwise isogenic, ErbB2-negative tumor variant.
Activation of ErbB2-specific T cells
To analyze the nature of the immune responses induced by C-ErbB2 and C-E-ErbB2 protein vaccines, BALB/c mice (five animals per group) were immunized twice by an i.p. injection of 0.5 nmol of the recombinant molecules at days 0 and 14 as described above. Five days after the last vaccination, the animals were sacrificed, and splenocytes were isolated and restimulated ex vivo with ErbB2-derived peptide TYLPTNASL (30). Activated CD8+ T cells were identified by flow cytometry after double-staining of splenocytes with Abs detecting CD8 and intracellular IFN-. Representative data from one animal of each group are shown in Fig. 5A. Mean values of the absolute numbers of CD8+IFN-+ cells from all five animals of each group are shown in Fig. 5B. Splenocytes from mice vaccinated with C-ErbB2 or C-E-ErbB2 proteins contained populations of activated CD8+ T cells that displayed markedly increased IFN- production upon restimulation with the ErbB2 peptide, indicating that immunization with both chimeric CTLA-4-ErbB2 molecules induced activation of ErbB2-specific CD8+ T cells. In contrast, in vitro restimulation of splenocytes from control animals with the ErbB2 peptide did not result in an increase in CD8+IFN-+ cells.
FIGURE 5. Immunization with CTLA-4-ErbB2 proteins induces ErbB2-specific CD8+ T cells. BALB/c mice were immunized by an i.p. injection of C-ErbB2 or C-E-ErbB2 proteins on days 0 and 14. Control animals received PBS. Five days later, splenocytes were isolated and stimulated for 5 h with H-2Kd-restricted ErbB2 peptide TYLPTNASL or left untreated before flow cytometric analysis with anti-CD8 and anti-IFN- Abs. A, Increase in CD8+IFN-+ cells upon peptide stimulation of splenocytes. Representative results from one animal of each group upon restimulation in the absence (upper panels) or presence of ErbB2 peptide are shown. B, Absolute numbers of CD8+ IFN-+ splenocytes (mean values from five animals per group).
Potential induction of Ab responses by CTLA-4-ErbB2 vaccines was also investigated. Thereby only low levels of ErbB2-specific serum Abs were detected in vaccinated animals 3 wk after immunization, making it unlikely that these contribute to tumor rejection (data not shown).
To determine the contribution of individual T cell subsets to the rejection of ErbB2-expressing tumor cells in immunized animals, BALB/c mice were vaccinated with C-ErbB2 and C-E-ErbB2 proteins on days 0 and 14 as described above, followed by s.c. challenge with Renca-lacZ/ErbB2 cells into each flank on day 16. Before tumor cell inoculation, CD8+ and/or CD4+ T cell populations were depleted by injection of CD8- and/or CD4-specific Abs on day 12. Injection of Abs was repeated on days 17 and 22. Successful depletion of T cells was controlled for each animal by FACS analysis (data not shown). Tumor volumes determined at day 34 of the experiment are shown in Fig. 6. In vaccinated mice depleted of CD8+ or CD8+ and CD4+ cells, tumors grew to sizes comparable to those of the PBS control group. In contrast, no tumors developed in animals vaccinated with C-ErbB2 or C-E-ErbB2 followed by injection of control Ab or Ab-depleting CD4+ cells. These results demonstrate that CD8+ effector cells induced by protein vaccination are absolutely required for tumor rejection.
FIGURE 6. Contribution of individual T cell subsets to the rejection of ErbB2-expressing tumor cells. BALB/c mice were injected with C-ErbB2 or C-E-ErbB2 proteins or PBS as described in the legend of Fig. 5. CD4+ and/or CD8+ T cells were depleted with anti-CD8 and/or anti-CD4 Abs on days 12, 17, and 22 as indicated. Control animals received rat IgG. Success of depletion was controlled for each mouse by FACS analysis (data not shown). On day 16, mice were inoculated with Renca-lacZ/ErbB2 tumor cells by a s.c. injection, and tumor growth was followed. Mean tumor volumes at day 34 of the experiment are shown.
Therapeutic vaccination
The influence of CTLA-4-ErbB2 protein vaccines on the growth of established tumors was investigated in animals carrying Renca-lacZ/ErbB2 tumors. BALB/c mice were inoculated with tumor cells by a s.c. injection (five animals per group). On day 2, the mice were treated by a s.c. injection of 0.5 nmol of C-ErbB2, C-E-ErbB2, or ErbB2222 proteins into the vicinity of each tumor. Control animals received PBS. Treatment was repeated on days 4 and 6. For protective vaccination, an i.p. injection of the vaccines was as efficient as a s.c. injection (data not shown). Nevertheless, because of the rapid growth of Renca-lacZ/ErbB2 tumors, for therapeutic vaccination experiments we chose s.c. injection of the proteins close to the tumors to better support immune responses in the tumor vicinity. Kinetics of tumor growth until day 43 of the experiment are shown in Fig. 7A. The survival graph in Fig. 7B displays the course of the disease in the animals, which were sacrificed latest when tumor size reached 0.8 cm3. All animals in the PBS control group had to be sacrificed due to tumor size latest until day 60 of the experiment (Fig. 7). In most vaccinated mice, tumor growth continued for a few days after onset of therapy before therapeutic effects became apparent (Fig. 7A). Forty days after tumor cell inoculation, all animals that were treated with C-ErbB2 had rejected both s.c. tumors, and the mice remained tumor free until this part of the experiment was terminated on day 96 (Fig. 7B). Therapy with C-E-ErbB2 protein initially resulted in tumor rejection in two of five animals. However, in one of these mice, tumors started to regrow on day 60, and the animal had to be sacrificed on day 73 (Fig. 7B). In the remaining mice of this group, delayed tumor growth was observed. Although injection of ErbB2222 control protein resulted in delayed tumor growth in some of the animals, none of these mice was cured (Fig. 7B). Similar results were obtained in a second experiment with smaller groups, where two of three animals treated with C-ErbB2 and three of three animals treated with C-E-ErbB2 rejected both tumors, but none of the three control mice treated with PBS. When the mice cured by C-ErbB2 or C-E-ErbB2 injection were i.v. rechallenged with Renca-lacZ/ErbB2 cells 3 mo after initial immunization (day 96), the animals in contrast to naive control mice remained completely free of pulmonary metastasis (data not shown).
FIGURE 7. Treatment with CTLA-4-ErbB2 proteins results in cures of tumor-bearing animals. BALB/c mice were inoculated with Renca-lacZ/ErbB2 tumor cells by a s.c. injection into each flank. When tumors were palpable, animals were treated by a s.c. injection of purified C-ErbB2 (?), C-E-ErbB2 (), or ErbB2222 () into the vicinity of each tumor as indicated. Treatment was repeated on days 4 and 6. Control animals received PBS (). A, Kinetics of tumor growth. Tumor growth was followed by caliper measurements, and mean tumor volumes were calculated. B, Course of the disease in mice from the experiment shown in A. Tumor-bearing mice were sacrificed latest when tumor size reached 0.8 cm3 to avoid suffering of the animals.
These results demonstrate that chimeric CTLA-4-ErbB2 proteins are not only effective as protective vaccines administered to healthy animals but are also active in a therapeutic setting inducing tumor regression and long-lasting immunity in tumor-bearing animals.
Discussion
Cancer immunotherapy has evolved as a promising strategy to use Ags differentially expressed on tumor cells for directed therapy. However, although potentially more effective than mAbs, specific cancer vaccines intended to induce endogenous antitumor responses still need to be optimized for successful clinical use (32, 33). Because induction of tumor-specific T cells is critical for cell-mediated immunity against tumors, targeting of Ags to professional APCs that can deliver effective costimulatory signals to T cells might overcome some of the limitations of present vaccination approaches.
In the present study, we have investigated whether fusion proteins designed to target a 222-aa fragment of the tumor-associated ErbB2 Ag to APCs in vivo can induce ErbB2-specific, cell-mediated antitumor immunity. The ErbB2222 fragment contains a number of CTL and Th epitopes previously shown to bind to human MHC class I and class II alleles (12, 34, 35, 36). It also includes the peptide epitope TYLPTNASL, which can be presented by human HLA-A2402 (37) and by H-2Kd in BALB/c mice (30). ErbB2 has been well recognized as a target for active immunotherapy, and a number of different strategies for vaccination have been investigated (38). Several studies have demonstrated effectiveness of ErbB2 DNA vaccines in murine models (9, 18, 19, 20). However, protein-based vaccination approaches, similar to the results obtained here with unmodified ErbB2222 protein, have only rarely achieved full protection of immunized animals (9, 23, 24, 25). Therefore, to enhance effectiveness of the ErbB2222 protein vaccine, we used the extracellular domain of human CTLA-4 for specific targeting to APCs. As a natural receptor, CTLA-4 binds with high affinity to the costimulatory B7 molecules CD80 and CD86, which are expressed predominantly on APCs (39, 40). Likewise, in the present study the bacterially expressed chimeric CTLA-4-ErbB2 proteins interacted specifically with CD80 and CD86 on the cell surface, whereas no binding to the cells was observed for ErbB2222 protein lacking a CTLA-4 domain. Upon binding, the proteins were internalized by B7-expressing Raji cells used as a model and degraded in intracellular compartments, a prerequisite for the generation of peptide epitopes and Ag presentation. CTLA-4-ErbB2 protein injected into the footpad of mice localized to areas with MHC II-expressing cells in the draining lymph node, making it likely that specific targeting of the chimeric molecules to APCs was retained in vivo.
Using respective DNA vaccine constructs in murine models, previously it was demonstrated that fusion of a CTLA-4 domain to model Ags such as human IgG or influenza virus hemagglutinin does not inhibit but can enhance Ag-specific humoral and cellular immune responses (41, 42). Nevertheless, so far it remained unclear whether such CTLA-4 fusions can also induce CTLs and cure established tumors (43). In the present study, vaccination of mice with chimeric CTLA-4-ErbB2 proteins induced ErbB2-specific CD8+ T cells recognizing the immunodominant TYLPTNASL epitope. Although cytotoxicity of such cells was not formally investigated, the failure of vaccinated mice to reject the challenge of ErbB2-expressing tumor cells upon depletion of CD8+, but not upon depletion of CD4+ cells, strongly suggests that cytotoxic CD8+ T cells are the main effector population operative in the elimination of tumor cells in vivo. Because mice were vaccinated with complete fusion proteins, in vivo activation of peptide-specific T cells must have been preceded by cellular uptake of the protein vaccines and suitable processing for presentation by MHC class I.
The immune responses induced by CTLA-4-ErbB2 protein vaccines were directed specifically against ErbB2-expressing tumor cells. Vaccinated mice were protected against the initial challenge with Renca-lacZ/ErbB2 but not ErbB2-negative Renca-lacZ cells, which still express the -galactosidase Ag. This indicates that the potent antitumoral responses observed were not caused by broad, unspecific immunostimulatory activity of the vaccines because of LPS contamination of the proteins or disturbance of endogenous B7-CTLA-4 interactions. In line with this, a similar CTLA-4-NY-ESO-1 construct failed to protect mice from challenge with Renca-lacZ/ErbB2 cells but induced rejection of Renca-lacZ cells expressing the relevant NY-ESO-1 tumor Ag (unpublished data). Vaccination with CTLA-4-ErbB2 proteins and tumor rejection induced immunological memory, resulting in long-term systemic immunity and protection against subsequent rechallenge with i.v. injected Renca-lacZ/ErbB2 cells several months later. Importantly, after vaccination with CTLA-4-ErbB2 proteins and rejection of initial challenge with ErbB2-expressing cells, mice now also acquired immunity to the ErbB2-negative tumor cell variant and rejected systemic challenge of Renca-lacZ cells. Possibly, tumor cell destruction by ErbB2-specific T cells resulted in the induction or enhancement of immune responses directed against additional Ags such as the bacterial -galactosidase, which in naive animals in our BALB/c-based model failed to serve as a tumor rejection Ag. In a clinical setting, similar determinant spreading could be beneficial to achieve effective immunity also against variant tumor cells that have lost expression of the Ag used for vaccination (44).
The C-E-ErbB2 molecule, in addition to the B7-specific cell binding domain and the antigenic ErbB2 fragment, carries the translocation domain of Pseudomonas ETA as an endosome escape function (29). After binding of full-length toxin to target cells and receptor-mediated endocytosis, this domain facilitates translocation of an enzymatically active ETA fragment to the cytoplasm (45). Previously, fusion of influenza virus peptide epitopes to an enzymatically inactive ETA derivative has been used to facilitate cellular uptake and enhance MHC I presentation of the Ags (46). More recently, vaccination of mice with a DNA vaccine encoding a fusion of ETA translocation domain with human papillomavirus type 16 E7 was reported to result in a 30-fold increase in E7-specific CD8+ T cells in comparison to unmodified E7 DNA vaccine (47). In these studies, ETA fragment and Ag were not connected to an APC-targeting domain. In the present study, the C-E-ErbB2 protein and C-ErbB2 lacking the ETA domain were very similar in their efficiency to induce ErbB2-specific CD8+ T cells and protect mice from tumor challenge. This suggests that specific targeting to APCs might already be sufficient to enhance cross-presentation or delivery of target Ag to an intracellular compartment suitable for MHC I loading (48), making a separate endosome escape function dispensable.
Vaccination of BALB/c mice with DNA vaccines encoding extracellular and transmembrane domains of the rat ErbB2 homologue Neu was reported previously to protect against lethal challenge with Neu-expressing tumor cells. Thereby CD4+ T cells and granulocytes were shown to be crucial for the antitumoral response, whereas depletion of CD8+ cells only partly affected protection (19). In contrast, while most likely required during initial priming of the response (49), during the effector phase, CD4+ cells were dispensable in animals immunized with C-ErbB2 or C-E-ErbB2 proteins. However, CD8+ T cells were absolutely essential for tumor rejection.
In this study concerned with the initial characterization of chimeric CTLA-4-ErbB2 protein vaccines, we have used human ErbB2 as an Ag in a murine model. Previously, CTLs from BALB/c mice that, similar to the CD8+ T cells identified in our study, recognize the human ErbB2 epitope TYLPTNASL have been shown to also recognize and lyse cells presenting the corresponding murine ErbB2 peptide TYLPANASL (30). This indicates that cross-reactive immune responses are possible and suggests that at least for this particular epitope tolerance to endogenous murine ErbB2 might have to be overcome to mount an effective response against the human Ag. However, in this context, it is noteworthy that in our BALB/c-based model also ErbB2222 protein lacking the CTLA-4 domain induced long-term immunity upon protective vaccination, albeit in a smaller number of animals. In contrast, in a therapeutic setting, ErbB2222 treatment only resulted in delayed tumor growth in some of the mice but was unable to cure any of the animals. To further investigate the possible influence of tolerance to the chosen Ag on the efficacy of such nontargeted and APC-targeted vaccines, subsequent studies will be performed in transgenic animals such as mice transgenic for human ErbB2 (20) or oncogenically activated or normal rat Neu (50, 51). Experiments with CTLA-4-Neu constructs in BALB-neuT mice (50) have been initiated.
Taken together, our results demonstrate that bacterially expressed chimeric proteins combining a tumor Ag and specific recognition of APCs in a single molecule can induce cell-mediated, Ag-specific immune responses and tumor rejection. Although in our model recombinant ErbB2222 on its own resulted in protective immunity in some of the vaccinated mice, it was ineffective as a therapeutic reagent and could not cure tumor-bearing animals. Fusion of this protein fragment to CTLA-4 strongly enhanced its efficacy, now also resulting in rejection of established tumor grafts.
Acknowledgments
We thank Manuela St?ber for purification of mAbs and Drs. Elke J?ger and Torsten Tonn for helpful suggestions and discussions.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported in part by Wilhelm Sander-Stiftung Grant 2002.010.1.
2 Address correspondence and reprint requests to Dr. Winfried S. Wels, Chemotherapeutisches Forschungsinstitut, Georg-Speyer-Haus, Paul-Ehrlich-Stra?e 42-44, D-60596 Frankfurt am Main, Germany. E-mail address: wels@em.uni-frankfurt.de
3 Abbreviations used in this paper: ETA, exotoxin A; RT, room temperature; X-Gal, 5-bromo-4-chloro-3-indolyl -D-galactoside.
Received for publication November 19, 2004. Accepted for publication February 8, 2005.
References
Van Der Bruggen, P., Y. Zhang, P. Chaux, V. Stroobant, C. Panichelli, E. S. Schultz, J. Chapiro, B. J. Van Den Eynde, F. Brasseur, T. Boon. 2002. Tumor-specific shared antigenic peptides recognized by human T cells. Immunol. Rev. 188:51.
Scanlan, M. J., A. O. Gure, A. A. Jungbluth, L. J. Old, Y. T. Chen. 2002. Cancer/testis antigens: an expanding family of targets for cancer immunotherapy. Immunol. Rev. 188:22.
Slamon, D. J., W. Godolphin, L. A. Jones, J. A. Holt, S. G. Wong, D. E. Keith, W. J. Levin, S. G. Stuart, J. Udove, A. Ullrich, et al 1989. Studies of the HER-2/neu proto-oncogene in human breast and ovarian cancer. Science 244:707.
Klapper, L. N., M. H. Kirschbaum, M. Sela, Y. Yarden. 2000. Biochemical and clinical implications of the ErbB/HER signaling network of growth factor receptors. Adv. Cancer Res. 77:25.
Olayioye, M. A., R. M. Neve, H. A. Lane, N. E. Hynes. 2000. The ErbB signaling network: receptor heterodimerization in development and cancer. EMBO J. 19:3159.
Slamon, D. J., B. Leyland-Jones, S. Shak, H. Fuchs, V. Paton, A. Bajamonde, T. Fleming, W. Eiermann, J. Wolter, M. Pegram, J. Baselga, L. Norton. 2001. Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N. Engl. J. Med. 344:783.
Rohrbach, F., B. Gerstmayer, M. Biburger, W. Wels. 2000. Construction and characterization of bispecific costimulatory molecules containing a minimized CD86 (B7-2) domain and single chain antibody fragments for tumor targeting. Clin. Cancer Res. 6:4314.
Azemar, M., S. Djahansouzi, E. Jager, C. Solbach, M. Schmidt, A. B. Maurer, K. Mross, C. Unger, G. von Minckwitz, P. Dall, B. Groner, W. S. Wels. 2003. Regression of cutaneous tumor lesions in patients intratumorally injected with a recombinant single-chain antibody-toxin targeted to ErbB2/HER2. Breast Cancer Res. Treat. 82:155.
Foy, T. M., J. Bannink, R. A. Sutherland, P. D. McNeill, G. G. Moulton, J. Smith, M. A. Cheever, K. Grabstein. 2001. Vaccination with Her-2/neu DNA or protein subunits protects against growth of a Her-2/neu-expressing murine tumor. Vaccine 19:2598.
Fisk, B., T. L. Blevins, J. T. Wharton, C. G. Ioannides. 1995. Identification of an immunodominant peptide of HER-2/neu protooncogene recognized by ovarian tumor-specific cytotoxic T lymphocyte lines. J. Exp. Med. 181:2109.
Knutson, K. L., K. Schiffman, M. A. Cheever, M. L. Disis. 2002. Immunization of cancer patients with a HER-2/neu, HLA-A2 peptide, p369-377, results in short-lived peptide-specific immunity. Clin. Cancer Res. 8:1014.
Disis, M. L., E. Calenoff, G. McLaughlin, A. E. Murphy, W. Chen, B. Groner, M. Jeschke, N. Lydon, E. McGlynn, R. B. Livingston, et al 1994. Existent T-cell and antibody immunity to HER-2/neu protein in patients with breast cancer. Cancer Res. 54:16.
Disis, M. L., S. M. Pupa, J. R. Gralow, R. Dittadi, S. Menard, M. A. Cheever. 1997. High-titer HER-2/neu protein-specific antibody can be detected in patients with early-stage breast cancer. J. Clin. Oncol. 15:3363.
Ward, R. L., N. J. Hawkins, D. Coomber, M. L. Disis. 1999. Antibody immunity to the HER-2/neu oncogenic protein in patients with colorectal cancer. Hum. Immunol. 60:510.
McNeel, D. G., L. D. Nguyen, B. E. Storer, R. Vessella, P. H. Lange, M. L. Disis. 2000. Antibody immunity to prostate cancer associated antigens can be detected in the serum of patients with prostate cancer. J. Urol. 164:1825
Knutson, K. L., K. Schiffman, M. L. Disis. 2001. Immunization with a HER-2/neu helper peptide vaccine generates HER-2/neu CD8 T-cell immunity in cancer patients. J. Clin. Invest. 107:477.
Disis, M. L., T. A. Gooley, K. Rinn, D. Davis, M. Piepkorn, M. A. Cheever, K. L. Knutson, K. Schiffman. 2002. Generation of T-cell immunity to the HER-2/neu protein after active immunization with HER-2/neu peptide-based vaccines. J. Clin. Oncol. 20:2624.
Pupa, S. M., A. M. Invernizzi, S. Forti, E. Di Carlo, P. Musiani, P. Nanni, P. L. Lollini, R. Meazza, S. Ferrini, S. Menard. 2001.
Prevention of spontaneous neu-expressing mammary tumor development in mice transgenic for rat proto-neu by DNA vaccination. Gene Ther. 8:75.
Curcio, C., E. Di Carlo, R. Clynes, M. J. Smyth, K. Boggio, E. Quaglino, M. Spadaro, M. P. Colombo, A. Amici, P. L. Lollini, P. Musiani, G. Forni. 2003. Nonredundant roles of antibody, cytokines, and perforin in the eradication of established Her-2/neu carcinomas. J. Clin. Invest. 111:1161
Piechocki, M. P., Y. S. Ho, S. Pilon, W. Z. Wei. 2003. Human ErbB-2 (Her-2) transgenic mice: a model system for testing Her-2 based vaccines. J. Immunol. 171:5787.
Rovero, S., K. Boggio, E. D. Carlo, A. Amici, E. Quaglino, P. Porcedda, P. Musiani, G. Forni. 2001. Insertion of the DNA for the 163-171 peptide of IL-1 enables a DNA vaccine encoding p185neu to inhibit mammary carcinogenesis in Her-2/neu transgenic BALB/c mice. Gene Ther. 8:447.
Cappello, P., F. Triebel, M. Iezzi, C. Caorsi, E. Quaglino, P. L. Lollini, A. Amici, E. Di Carlo, P. Musiani, M. Giovarelli, G. Forni. 2003. LAG-3 enables DNA vaccination to persistently prevent mammary carcinogenesis in HER-2/neu transgenic BALB/c mice. Cancer Res. 63:2518(Florian Rohrbach, Robert )