Targeting of the non-mutated tumor antigen HER2/neu to mature dendritic cells induces an integrated immune response that protects against breast cancer in mice

Animal experiments were designed to fulfill the ethical and scientific principles provided by the Institutional Animal Care and Use Committee of The Rockefeller University (New York, NY, USA) (approved protocol 08117). Mice were maintained under specific pathogen-free conditions and were 6 to 8 weeks of age. C57BL/6, BALB/c, FVB/N, and HLA-A2.1 transgenic mice in the C57BL/6 background - C57BL/6-Tg(HLA-A2.1)1Enge/J - were purchased from The Jackson Laboratory (Bar Harbor, ME, USA). DEC^-/- mice were generated and provided by Michel Nussenzweig (The Rockefeller University) and are available from The Jackson Laboratory. At least three mice per group were used in immunization experiments.

The neu-expressing mammary tumor cell line NT2.5 was derived from a spontaneous mammary tumor in female neu-N mice (FVB/N background). The cell line was established and kindly provided by Elizabeth M Jaffee (Johns Hopkins University School of Medicine, Baltimore, MD, USA). NT2.5 tumor cells were grown in a previously defined breast media, which consisted of RPMI (Gibco, now part of Invitrogen Corporation, Carlsbad, CA, USA) with 20% fetal bovine serum, 1% L-glutamine, 1% non-essential amino acids, 1% Na pyruvate, 0.5% penicillin/streptomycin, 0.02% gentamicin (Invitrogen Corporation), and 0.2% insulin (Sigma-Aldrich, St. Louis, MO, USA). Cells were maintained at 37°C in 5% CO₂. The HER2 stably transfected tumor cell line, E0771/E2, was generously provided by Wei-Zen Wei (Karmanos Cancer Institute, Wayne State University, Detroit, MI, USA). Anti-CD4 (clone GK1.5) and anti-CD8 (clone 2.43) hybridoma cells were obtained from the American Type Culture Collection (Manassas, VA, USA) and maintained in accordance with its protocols.

DNA coding HER2 extracellular domain (amino acid 22-653) was cloned in frame into the COOH terminus of anti-DEC (DEC-HER2) or control IgG heavy chain (Ctrl Ig-HER2) as described previously [30]. Fusion mAb was expressed by transient transfection in 293T cells and purified on protein G columns (GE Healthcare Bio-Sciences Corp., Piscataway, NJ, USA). Purified mAb was characterized by SDS-PAGE and Western blot by using anti-mouse IgG- horseradish peroxidase (IgG-HRP) (SouthernBiotech, Birmingham, AL, USA) or anti-HER2 mAb (clone 42; BD Transduction Laboratories, San Jose, CA, USA). Specific binding of the fusion mAb was verified by using Chinese hamster ovary (CHO) cells stably transfected with mouse DEC receptor. Binding was detected by flow cytometry by using phycoerythrin (PE)-conjugated goat anti-mouse IgG mAb (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA) and Alexa Fluor 488-conjugated mouse anti-HER2 mAb (clone 24D2; BioLegend, San Diego, CA, USA). All Abs had less than 0.125 endotoxin units per milligram in a Limulus Amebocyte Lysate assay (QCL-1000; BioWhittaker, Walkersville, MD, USA).

Overlapping (staggered by four amino acids) 15-mer peptides covering the HER2 and neu extracellular domain and HIV gag p24 protein were synthesized by Henry Zebroski in the Proteomics Resource Center of The Rockefeller University. The use of peptides overcomes, in large part, the need for antigen processing by APCs during the immune assays. The 161- and 147-member HER2 and neu peptide libraries were divided into seven and six pools, respectively.

Mice were immunized intraperitoneally with 5 μg of DEC-HER2 or Ctrl Ig-HER2 fusion mAb in combination with 50 μg of poly IC (polyinosinic/polycytidylic acid) (InvivoGen, San Diego, CA, USA). When indicated, a combination of 50 μg of poly IC and 25 μg of agonistic anti-CD40 mAb (clone 1C10) was used to mature DCs.

Bulk splenocytes were stimulated with specific peptide pools (2 μg/mL or indicated concentration) or medium alone in the presence of a co-stimulatory anti-CD28 mAb (clone 37.51) for 6 hours. Brefeldin A (10 μg/mL) (Sigma-Aldrich) was added for the last 5 hours to accumulate intracellular cytokines. Anti-CD28 mAb was used only in a 6-hour intracellular cytokine staining assay but not in other T-cell immune assays, including enzyme-linked immunosorbent assay (ELISA), enzyme-linked immunosorbent spot (ELISPOT), and CFSE (5,6-carboxy fluorescein diacetate succinimidyl ester) dilution assays. For functional avidity analysis, graded doses (10 to 0.0016 μg/mL) of peptides were used to re-stimulate splenocytes. After stimulation, cells were washed and then incubated with anti-CD16/CD23 mAb (clone 2.4G2) to block Fcγ receptor for 15 minutes at 4°C. Cells were stained with Live/Dead Fixable Aqua vitality dye (Invitrogen Corporation), fluorescein isothiocyanate-conjugated anti-CD4 (clone RM4-5), PerCP-Cy5.5-conjugated anti-CD8 (clone 53-6.7), and Pacific blue-conjugated anti-CD3 (clone 17A2) (eBioscience, San Diego, CA, USA) for 20 minutes at 4°C. Cells were fixed, permeabilized (Cytofix/Cytoperm Plus; BD Biosciences, San Jose, CA, USA), and stained with allophycocyanin-conjugated anti-interferon-gamma (anti-IFNγ), PE-conjugated anti-IL-2, and PECy7-conjugated anti-tumor necrosis factor-alpha (anti-TNF-α) mAbs for 20 minutes at 4°C (BD Biosciences) and resuspended in stabilizing fixative (BD Biosciences). Data were collected by using a BD LSR II flow cytometer (BD Biosciences) and analyzed with FlowJo software (Tree Star, Inc., Ashland, OR, USA).

A CFSE dilution assay was used to assess the proliferative capacity of T cells. Bulk splenocytes (2 × 10⁷ cells/mL) were labeled with 2.5 μM CFSE (Invitrogen Corporation) in a 37°C water bath for 10 minutes. CFSE-labeled T cells were re-stimulated with pools of peptide (0.2 μg/mL) for 4 days, often in combination with intracellular cytokine staining of cells re-stimulated for the last 6 hours of culture.

Multi-screen-HA MAHA 54510 (Millipore, Billerica, MA, USA) plates were coated with 10 μg/mL of purified rat anti-mouse-IFNγ mAb (clone R46A2; BD Biosciences) in phosphate-buffered saline (PBS) overnight at 4°C. Plates were washed and blocked with PBS/1% bovine serum albumin (BSA) for 1 hour at 37°C. Magnetic-activated cell sorting (MACS)-purified CD8⁺ or CD4⁺ T cells (3 × 10⁵) were cultured for 2 days with 1 × 10⁵ purified CD11c⁺ spleen DCs pulsed with the peptide mix (1 μg/mL) or NT2.5 tumor lysate (10 μg/mL). Biotin-conjugated rat anti-mouse-IFNγ mAb (clone XMG 1.2, 2 μg/mL; BD Biosciences) was used as the detection Ab. After 2-hour incubation with detection Ab, spots were visualized with a Vectastain ABC kit (Vector Laboratories, Inc., Burlingame, CA, USA), followed by diaminobenzidine as the substrate (Invitrogen Corporation). Spots were counted in an ELISPOT reader (Autoimmun Diagnostika GmbH, Straβberg, Germany).

Splenic CD4⁺ and CD11c⁺ cells were purified by MACS. CD4⁺ cells (3 × 10⁵) were incubated with 1 × 10⁵ CD11c⁺ cells with peptide mix (2 μg/mL) in 96-well U-bottomed plates for 48 hours. Concentrations of IFNγ, IL-4, IL-10, and IL-17 in supernatant were measured by Ready-Set-Go! ELISA sets (eBioscience).

To detect HER2-specific Ab response, we produced FLAG-HER2 soluble protein by transient transfection of 293T cells and purification with anti-FLAG affinity gel (Sigma-Aldrich). The quality of FLAG-HER2 protein was verified by SDS-PAGE gel under non-reducing conditions (Figure S1 of Additional file 1). We coated high-binding ELISA plates (Nunc; Thermo Fisher Scientific Inc., Rochester, NY, USA) with 500 ng/mL (50 ng/well) of FLAG-HER2 overnight at 4°C. Plates were washed with PBS/0.1% Tween-20 and blocked with PBS/0.1% Tween 20/5% BSA for 1 hour at 37°C. Serial dilutions of serum were added to the plates and incubated for 1 hour at 37°C. Secondary goat anti-mouse IgG-specific Abs conjugated with HRP (SouthernBiotech) were added and visualized with tetramethylbenzidine (eBioscience) at room temperature for 5 to 10 minutes. To determine the IgG isotype, anti-mouse IgG1 or IgG2a Abs were used. The reported titers represent the highest dilution of sample showing an OD₄₅₀ (optical density at 450 nm) of higher than 0.1. The data were presented as the log₁₀ Ab titer. To determine whether serum IgG can bind to HER2/neu-expressing tumor cells, serial diluted serum was incubated with E0771/E2 (HER2⁺) or NT2.5 (neu⁺) tumor cells for 15 minutes at 4°C. Anti-mouse IgG-PE mAbs were used to detect the binding of serum IgG to tumor cells. Data were acquired by using a BD LSR II flow cytometer.

FVB/N mice were immunized intraperitoneally with 5 μg of DEC-HER2 or Ctrl Ig-HER2 mAb together with 50 μg of poly IC on days 0 and 28. Poly IC alone (50 μg) was injected as a negative control. Ten days after the boost immunization, mice were inoculated subcutaneously with 1 × 10⁶ NT2.5 tumor cells in the shaved right flank. Tumor size was measured three times every week by using a caliper. Tumor volumes were estimated according to the formula: length × (width)² × 0.5. For survival analysis, tumor sizes of at least 500 mm³ were defined as the experimental endpoint. For Ab depletion, 200 μg of CD4 or CD8 mAbs or both were given to mice intraperitoneally after boost immunization 9, 6, and 3 days before tumor challenge. Isotype control rat IgG was given as a negative control. Efficiency of depletion was confirmed by fluorescence-activated cell sorting (FACS) analysis of peripheral blood cells.

All analysis was performed by using Prism 4.0 GraphPad software (GraphPad Software, Inc., San Diego, CA, USA). A two-sided Student t test (between two groups or conditions) was applied to compare statistical significance between peptide-specific responses and treatment groups of immunized mice. Survival studies were analyzed by Kaplan-Meier survival curves and log-rank test. Results were considered statistically significant when the P value was less than 0.05.

To deliver HER2 protein to mouse DCs directly in vivo, we cloned the extracellular domain of HER2 (amino acid 22-653) in frame into the heavy chain of the anti-mouse DEC mAb (DEC-HER2) (Figure 1A). The heavy chain of an isotype-matched non-reactive control IgG was also engineered as a non-targeting control (Ctrl Ig-HER2). The fusion mAbs were composed of a 140- to approximately 150-kDa heavy chain, consistent with a predicted mass of 90 kDa for the HER2 extracellular domain fused to the approximately 50-kDa heavy chain of unconjugated mouse IgG1 (Figure 1B). To verify whether the fusion mAb bound to the mouse DEC receptor, a stable CHO cell transfectant, expressing mouse DEC on the surface, was stained with the indicated concentration of mAb. As analyzed by FACS, DEC-HER2 mAb, but not Ctrl Ig-HER2 mAb, demonstrated proper DEC-binding activity (Figure 1C).

To determine the immunogenicity of the DEC-HER2 fusion mAb, we immunized C57BL/6 (H-2^b) mice with DEC-HER2 or Ctrl Ig-HER2 intraperitoneally at a dose of 5 μg (equivalent to 14 pmol or 2.7 μg of HER2 protein). Poly IC (50 μg) and agonistic anti-CD40 mAb (25 μg) were used to mature DCs. In C57BL/6 mice, CD4⁺ T cells responded strongly to epitopes present in HER2 peptide pools 4 and 5 (Figure 2A). We also found a weak response against HER2 epitopes in pools 3 and 7 (Figure 2A). When CD28 mAb was removed from culture, T-cell responses were reduced but remained significantly higher than background signals (Figure S2A of Additional file 2). The T-cell response was specific to HER2 since there was no reactivity to HIV gag peptides. Although Ctrl Ig-HER2 vaccination induced a significant CD4⁺ T-cell response against peptide pool 5, its magnitude was significantly weaker than that induced by DEC-HER2 (Figure 2A). The breadth of the response was limited to pool 5. We did not observe significant CD4⁺ T-cell responses against other pools.

In addition to determining effector T-cell pool size, we determined the proliferative capacity of vaccine-induced CD4⁺ T cells upon antigen re-stimulation by CFSE dilution assays. Immunized T cells were stimulated with HER2 or control peptides in vitro for 4 days. When recalled with HER2 peptide pools for the last 6 hours, the proliferating CD4⁺ T cells produced IFNγ specifically against HER2 (Figure 2B).

To determine whether vaccine-induced CD4⁺ T cells are multi-functional (as defined by the combination of IFNγ, IL-2, and TNFα production at the single-cell level [36]), we measured the relative proportion of potential combinations of cytokines as depicted by pie charts and shown in Figure S3 (left panel) of Additional file 3. The mean percentage of CD4⁺ T cells producing IFNγ/IL-2/TNFα was approximately 57%. Thus, DEC-HER2 vaccination elicited multi-functional CD4⁺ T-cell responses.

To determine the types of Th responses induced by vaccination, we measured the production of the Th1/Th2/Th17 cytokines (IFNγ, IL-4, IL-10, and IL-17) by ELISA. CD4⁺ T cells from immunized mice were co-cultured with CD11c⁺ cells in the presence of HER2 peptide pool 5 or HIV gag peptide mix for 48 hours. As shown in Figure 2C, in mice immunized with DEC-HER2, we detected a dominant Th1 response with little Th2 or Th17 cytokine secretion from mice immunized with DEC-HER2. T cells from DEC-HER2 immunized mice produced significantly higher amounts of IFNγ but similar levels of IL-4/IL-10/IL-17 in comparison with T cells from mice immunized with Ctrl Ig-HER2 mAb. These data demonstrate that targeting DCs via DEC primed strong Th1 responses against HER2.

As a monotherapy, poly IC has been tested at high doses in patients with cancer and has been shown to have a favorable safety profile [37‐41]. To assess the adjuvant activity of poly IC, we primed and boosted mice with DEC-HER2 in combination with 50 μg of poly IC but without anti-CD40 mAb. We found that poly IC as the only adjuvant mediated significant CD4⁺ T-cell responses in mice immunized with DEC-HER2 mAb but not in mice treated with Ctrl Ig-HER2. We identified four reactive HER2 peptide pools, namely pools 3, 4, 5, and 7 (Figure 3A), which are the same immunogenic pools when poly IC and CD40 Ab were used together as the adjuvants. The mean proportion of CD4⁺ T cells producing IFNγ/IL-2/TNFα was approximately 50% (right panel of Figure S3 of Additional file 3), which is similar to the response when poly IC and CD40 Ab were co-administrated. Thus, DEC-HER2 vaccination with poly IC can elicit multi-functional CD4⁺ T-cell responses.

To our knowledge, HER2-specific CD4 T-cell epitopes have not been identified in C57BL/6 mice. To identify the individual reactive CD4 epitopes, we re-stimulated immune CD4⁺ T cells with single peptides from pools 3, 4, 5, and 7. Seven peptides were able to induce significant IFNγ production by CD4⁺ T cells from DEC-HER2 immunized mice (Figure S4 of Additional file 4 and Table 1). The dominant CD4 T-cell epitopes in C57BL/6 mice locate within the HER2_284-302 region (NPEGRYTFGASCVTACPYN).

To determine whether DEC-HER2 can induce CD4⁺ T-cell responses in different MHC backgrounds, we immunized BALB/c (H-2^d) and FVB/N (H-2^q) mice. Significant CD4⁺ T-cell response was induced with DEC-HER2 was targeted to the DEC receptor, although the magnitude was lower than that observed in C57BL/6 mice (Figure 3B, C). The dominant epitope or epitopes presented in BALB/c mice were located in pool 2 (Figure 3B). The amino acid sequence spanning HER2 peptide pool 2 is shown in Table S1 of Additional file 5, and the predicted I-A^d-restricted epitopes are shown in Table S2 of Additional file 6. CD4⁺ T-cell responses can also be induced in the FVB/N mice, and the dominant epitope or epitopes are located in HER2 peptide pool 4 (Table S1 of Additional file 5 and Figure 3C).

To determine whether the DEC receptor is essential for vaccination, we immunized DEC^-/- mice. DEC-HER2 did not induce HER2-specific T-cell responses in DEC^-/- mice (Figure 3D). The dependence of DEC expression in this model is consistent with our previous studies using other immune antigens [32, 42].

To determine the pattern recognition receptor required for poly IC, we tested TLR3^-/- and TLR3^-/-MDA5^-/- mice. Though significantly reduced, half of the adjuvant effect of poly IC remained intact in TLR3^-/- mice. In contrast, poly IC completely lost its adjuvant effect in TLR3^-/-MDA5^-/- mice (Figure 3E). These data suggest that both endosomal TLR3 and the cytosolic sensor MDA5 are required for the maximal adjuvant effect of poly IC.

Thus, targeting HER2 to DCs activated by poly IC induced potent, broad, and multi-functional CD4⁺ T-cell immunity in mice; these are important features of a successful vaccination. Induction of T-cell immunity was dependent on the expression of DEC receptor for antigen uptake and TLR3/MDA5 for DC maturation.

To assess cross-priming of HER2 protein by DEC targeting, we vaccinated FVB/N (H-2^q) mice with DEC-HER2 or Ctrl Ig-HER2 in combination with poly IC. We found that HER2-specific CD8⁺ T-cell responses were elicited in mice vaccinated with DEC-HER2 but not in mice vaccinated with Ctrl Ig-HER2 mAb (Figure 4A). The dominant epitope or epitopes are located in peptide pool 5 (Figure 4A), and the amino acid sequence spanning pool 5 is shown in Table S1 of Additional file 5. We also found that HER2-specific CD8⁺ T cells can proliferate and produce IFNγ when re-stimulated with HER2 peptides (200 ng/mL) (Figure 4B).

To test whether DEC targeting can enhance cross-priming in a human HLA haplotype, we administered DEC-HER2 + poly IC to HLA-A2 transgenic mice on the C57BL/6 background. Although we did not detect cross-priming in wild-type C57BL/6 mice, the HLA-A2 transgenic mice developed HER2-specific CD8⁺ T-cell responses, as analyzed by ELISPOT assay (Figure 4C). We identified that peptide pool 5 contained the reactive CD8 epitope or epitopes. The amino acid sequence spanning HER2 pool 5 is shown in Table S1 of Additional file 5. Two previously identified HLA-A2-restricted epitopes are located in peptide pool 5. They are HER2_435-443 (ILHNGAYSL) and HER2_466-474 (ALIHHNTHL) [43, 44]. Thus, targeting HER2 to DCs enhanced not only Th1 responses but also cross-presentation to CD8⁺ T cells.

To determine whether T cells induced by human HER2 protein vaccination are able to cross-react with the homologous rat neu protein, we re-stimulated T cells with neu peptides. Despite amino acid sequence differences between human HER2 and rat neu, we found that CD4⁺ T cells induced by DEC-HER2 vaccination were able to secrete IFNγ upon re-stimulation with the neu peptide pool (Figure 5A). To estimate the functional avidity of the vaccine-induced CD4⁺ T cells, we re-stimulated splenocytes from mice immunized with DEC-HER2 + poly IC with graded concentrations of HER2 peptide pool 5 or neu peptide pool 4 (from 10 to 0.0016 μg/mL or 6.67 to 0.001 μM). As shown in Figure 5B, HER2- and neu-specific T cells have a similar EC₅₀ (that is, concentration of peptide that leads to 50% of the maximal responses). Thus, xeno-priming with HER2 protein can induce homologous rat neu-specific CD4⁺ T-cell responses; this induction is important for tumor vaccination studies using rat neu-expressing tumor models.

To determine whether targeting HER2 to DEC can generate humoral responses, we measured serum anti-HER2 total IgG by ELISA. In FVB/N mice, both DEC-HER2 and Ctrl Ig-HER2 mAbs induced HER2-specific Ab responses with a similar titer of at least 1,800 in total IgG and at least 600 in IgG1 or IgG2a (Figure 6A).

To investigate whether the serum IgG can recognize natural HER2/neu protein presented on tumor cells, we performed flow cytometry-based assays. Serum IgG bound to the HER2-expressing tumor cell line E0771/E2 up to a 1:400 dilution (Figure 6B). Immune sera also bound to neu protein on the surface of NT2.5 tumor cells at a 1:100 dilution, although the binding was much weaker (approximately 10-fold) than binding to HER2 protein (Figure 6C). These data indicate that cross-reactivity of serum HER2-specific IgG to neu protein is weak, unlike what we observed with T-cell cross-reactivity (Figure 5). The baseline mean fluorescence index without immune serum was similar between the two tumor cell lines.

To assess whether the vaccine-induced HER2/neu-specific T-cell immunity can mediate protective anti-tumor immunity, we immunized FVB/N mice with DEC-HER2 or Ctrl Ig-HER2 in combination with poly IC. Ten days after boost immunization, mice were challenged with NT2.5 tumor cells in the mammary fat tissue. Vaccination with only 5 μg of DEC-HER2 protein with poly IC significantly delayed tumor growth (Figure 7A). Although Ctrl Ig-HER2 vaccination was able to induce anti-HER2 Ab responses (Figure 6), it was unable to protect mice from tumor outgrowth. As expected, administration of poly IC alone did not delay tumor growth. Although not all of the mice that received DEC-HER2 + poly IC vaccinations were tumor-free, the survival rate (up to 80 days) of DEC-HER2-immunized mice was significantly greater than that of Ctrl Ig-HER2-treated or untreated mice.

To determine the underlying tumor protection mechanism, we depleted CD4⁺, CD8⁺, or both populations at the effector phase, the time of tumor challenge. Depletion efficiency was verified by FACS analysis of peripheral blood lymphocytes the day before tumor challenge (Figure S5 of Additional file 7). We found that both CD4⁺ and CD8⁺ T cells were required for tumor protection in the vaccinated mice as determined by tumor growth kinetics (Figure 7C). Survival analysis (up to 60 days) shows that depletion of CD4⁺ T cells did not completely abrogate survival benefit of DEC-HER2 vaccination. However, depletion of CD8⁺ T cells completely abrogated the tumor protection effect (Figure 7D). These data suggest that CD8⁺ T cells play a more critical role for long-term survival than CD4⁺ T cells do.

To determine the induction of anti-tumor T-cell responses, we harvested the splenocytes 14 days after tumor challenge and re-stimulated purified CD4⁺ or CD8⁺ T cells with purified splenic CD11c⁺ cells pulsed with peptides mix (HIV gag, HER2, or neu) or NT2.5 tumor lysate. Strong HER2 and neu-specific CD4⁺ (Figure 7E) and CD8⁺ (Figure 7F) T-cell responses against HER2/neu were induced by DEC-HER2 but not by Ctrl Ig-HER2. CD4⁺ and CD8⁺ T cells also produced IFNγ upon re-stimulation with NT2.5 tumor lysate-pulsed DCs. Thus, DEC-HER2 xeno-priming significantly increased anti-tumor immunity and this increase led to a significant delay in tumor outgrowth. Thus, vaccine-induced CD4⁺ and CD8⁺ T cells were both essential for tumor protection and long-term survival.