Introduction
Breast cancer affects women worldwide and is a major public health problem. Despite progress in the field of surgery and adjuvant therapies, the risk of metastatic relapse remains high. Human Epidermal growth factor Receptor 2 (HER2) over-expression is observed in approximately 20% of invasive breast cancer and is an independent predictor of survival as it is associated with poor prognosis, aggressive disease and resistance to chemotherapy and hormone therapy [
1‐
4]. HER2 has been targeted with immunotherapeutic approaches based on the use of anti-HER2 monoclonal antibodies (mAb), tyrosine kinase inhibitors and cancer vaccines [
5].
Patients with HER2-expressing tumors show HER2-specific humoral and/or T-cell responses [
6,
7]. Such anti-HER2 immune responses, albeit of low magnitude, indicate that HER2 is a suitable candidate for HER2-targeted vaccine strategies. Induction of a stronger HER2-specific immunity with anti-tumor vaccines should lead to the establishment of immune memory, thereby preventing tumor recurrence, metastasis and relapse. However, HER2-induced immunological tolerance has been described, probably related to its oncofetal origin, which is an obstacle to efficient vaccination against this antigen [
8]. To circumvent self antigen-dependent tolerance, peptide-, DNA- or anti-Idiotype (Id)-based vaccines have been developed that show great specificity without notable toxicity [
9‐
14]. Among them, anti-Id antibodies have been proposed as vaccines for cancer immunotherapy and significant success has been achieved using anti-Id vaccines mimicking tumor-associated antigens (TAAs). This approach is based on N.K. Jerne's idiotype network theory about the Ab1-Ab2-Ab3 antibody cascade stimulation, whereby specific anti-Id antibodies (Ab2β induced by immunization with antigen-specific Ab1 antibodies, can serve as an "internal image" of the target antigen and can be used to induce Ab3 (also named Ab1') antibodies that can bind to the cognate antigen [
15]. Previous studies have described the use in solid tumors of anti-Id mAbs, which mimic TAAs, such as carcinoembryonic antigen (CEA), disialoganglioside GD2 or cancer-antigen 125 (CA-125), and demonstrated that these anti-Id mAbs induce an antigen-specific humoral response [
16‐
19]. In clinical trials, including patients with ovarian carcinoma, colorectal carcinoma or malignant melanoma, anti-Id-specific humoral and/or cellular responses following immunization were associated with a better survival rate without toxicity, but with modest objective responses. Available results of treatment of breast cancer patients with anti-Id mAbs are still very preliminary and conclusions go no further than the mere biologic proof of principle [
20].
In this context, our goal was to develop a vaccine to boost anti-HER2 immunity in patients with HER2-positive tumors and pre-existent low-level immunity. To this end, the use of HER2-mimicking anti-Id antibodies as a vaccine is a promising alternative. In a previous work [
21], we reported that two human anti-Id scFv antibody fragments (scFv40 and scFv69), which were selected by screening a phage-displayed library using the anti-HER2 antibody trastuzumab, induced an anti-HER2 antibody response in sera of immunized BALB/c mice. In the present study, we show that immunization with anti-Id scFv40 and scFv69 induces production of Ab1' that inhibit growth of HER2-positive tumor cells both
in vitro and
in vivo. Moreover, prophylactic vaccination with anti-Id scFv69 efficiently protects MMTV.f.huHER2(Fo5) mice from developing spontaneous HER2 positive mammary tumors through the induction of a HER2-specific Ab1' antibody response. Taken together, these results indicate that the anti-Id scFv69 fragment could be envisaged as an anti-idiotype-based vaccine for adjuvant therapy in patients with HER2-positive tumors, through reversion of HER2-specific immunological tolerance.
Materials and methods
Reagents
The recombinant HER2-Fc fusion protein (kindly provided by Pr. J.P. Mach, Biochemistry Institute, University of Lausanne, Switzerland) is composed of two extracellular domains of human HER2 linked with a human Fc fragment and was produced by transfection of Human Embryonic Kidney 293 cells. The anti-HER2 humanized antibody trastuzumab (Herceptin
®) was purchased from Genentech, Inc. (San Francisco, CA, USA). Production, purification and characterization of anti-Id trastuzumab-selected scFv40 and scFv69 have already been described [
21]. The irrelevant scFv 13R4 is a kind gift from Dr. P. Martineau (IRCM, Montpellier, France). MFE23-Fc fusion protein is a kind gift from R. Kontermann (University of Stuttgart, Germany).
Cell lines
The HER2-overexpressing human ovarian carcinoma cell line SK-OV-3 and the Chinese hamster ovarian (CHO) cell line, were obtained from the American Type Culture Collection (ATCC; Rockville, MD, USA). SK-OV-3 cells were cultured in DMEM medium (Gibco, Paisley, UK) and CHO cells in RPMI-1640 medium (Gibco), both supplemented as recommended by ATCC. Cells were maintained at 37°C in a humidified atmosphere with 5% CO2.
Mice
All
in vivo experiments were performed in compliance with the French guidelines for experimental animal studies (Agreement N° B34-172-27). Six-to eight-week-old female BALB/c, FVB (the mouse strain used to generate the MMTV.f.huHER2(Fo5) transgenic line) and nude athymic nude mice were purchased from Harlan Laboratories (Germany). The transgenic mouse line MMTV.f.huHER2(Fo5) was obtained from Genentech and has been previously described [
22]. These mice, which over-express human HER2 under the control of the MMTV promoter and spontaneously develop mammary tumors within an average time of seven months from birth, are used as a pre-clinical model of HER2-overexpressing breast cancer. They were maintained in the IRCM animal facilities and used for immunization when they were three-month-old. For
in vivo experiments, body weight was assessed weekly, as a surrogate marker of toxicity for the duration of the experiments, concomitantly with tumor measurement.
Anti-Id scFv mouse immunization
Immunogens were first injected subcutaneously (s.c.) after emulsion with Complete Freund Adjuvant (CFA) in BALB/c, FVB or MMTV.f.huHER2(Fo5) mice, followed two weeks later by a second s.c. administration with Incomplete Freund Adjuvant (IFA). Then, two intraperitoneal (i.p.) injections with IFA were given at day 21 and 35 after the initial boost. In FVB-MMTV.f.huHER2(Fo5) mice, a final i.p. injection with IFA was given at Day 90 after the initial boost. Immunogens were anti-Id scFv40 and scFv69 (50 μg/inj), HER2-Fc fusion protein (20 μg/inj), or phosphate-buffered saline PBS (negative control) at a 1:1 ratio with CFA or IFA. For serum antibody measurements, mice were bled and sera were drawn from the tail vein at various times during the experiment and stored at -20°C until assay.
ELISA and flow cytometry analysis of HER2-specific Ab1' antibodies in sera of vaccinated mice
Indirect ELISA and flow cytometry analysis were performed, as previously described [
21], to detect the presence of anti-HER2 antibodies (Ab1') in sera of immunized mice. Briefly, indirect ELISA was performed by coating 96-well plates overnight at 4°C with HER2-Fc fusion protein diluted at 5 μg.ml
-1 in PBS and then saturated for one hour with PBS containing 1% serum albumin (BSA). After four washings with 0.1% Tween/PBS, 100 μl of diluted sera (1:50) from immunized mice were added to each well and incubated for 1.5 h at room temperature. Secondary goat anti-mouse IgG (γ-chain specific)-horseradish conjugate (Millipore Chemicon, Billerica, MA, USA) was added (dilution 1:2,500) after plate washing, incubated for 1.5 h, and then
ο-phenylenediamine substrate was added for revelation according to the supplier's instructions (Sigma, St Louis, MO, USA). After stopping the reaction with 50 μl of 3N HCl/well, absorbance was measured at 490 nm, using an ELISA microreader (Multiskan EX, Thermo Electron Corporation, Vantaa, Finland).
For flow cytometry analysis, SK-OV-3 and CHO cells were incubated for one hour at 4°C with 100 μl of each mouse serum (diluted at 1:50) or 100 μl of trastuzumab (at 20 μg/ml) diluted in PBS containing 10% fetal calf serum. After washing, cells were incubated at 4°C with either 100 μl of sheep anti-mouse IgG-FITC-labeled antibody (1:100; Sigma-Aldrich, Saint Louis, MO, USA) for sera or anti-human-FITC-labeled antibody (1:100; Sigma-Aldrich) for trastuzumab for 45 minutes. Cells were then suspended in 500 μl PBS and 10,000 events were analyzed with a FACScan apparatus (Becton Dickinson, Franklin Lakes, NJ, USA).
Stable transfection of SK-OV-3 cells with luciferase
To image tumor cells in nude mice, SK-OV-3 cells were transfected with a luciferase-expressing vector (CMV-Luc-Hygro), a gift from Dr. P. Balaguer (U896, Institut de Recherche en Cancérologie de Montpellier, France), using a Calcium Phosphate transfection kit (Sigma). Transfected cells (SK-OV-3-Luc) were selected with 400 μg/ml Hygromycin and individual clones were analyzed for luciferase expression by measuring luciferase activity of cells plated in 96 well/plates with a luminometer (Microbeta, Perkin-Elmer Wallac, Waltham, MA, USA). HER2 expression in SK-OV-3-Luc clones was measured by flow cytometry with a FACScan flow cytometer (Becton Dickinson, San Jose, CA, USA).
Cell viability
Sera from immunized BALB/c mice were harvested at Day 42 after the initial boost, pooled, sterile-filtered and stored at -20°C until use. HER2-specific antibodies were detected in the serum as described above. A total of 1 × 104 SK-OV-3-Luc cells/well were plated in 96-well plates in 100 μl of culture medium and grown at 37°C for 24 h. Thereafter, 100 μl of a 1:5 dilution of sera from immunized BALB/c mice were added to the culture medium. On Day 5, 50 μl of luciferin solution (Promega), at a final concentration of 3 × 10-4 M, were added to the cells and luminescence was measured with a luminometer (Microbeta, Wallac). Cell number was proportional to the emitted luminescence. The results were normalized to the luminescence of non-treated cells. Each condition was done in triplicate in two independent experiments.
Adoptive transfer of immune sera from anti-Id scFv-vaccinated mice
Female athymic nude mice were i.p. injected with 5 × 105 SK-OV-3-Luc cells/mouse and grouped (n = 5 animals/group) at Day 3 after injection. At Day 4 post-graft, mice were i.p. injected with 200 μl of mouse serum from naïve donors, or from BALB/c mice immunized with anti-Id scFv40, anti-Id scFv69, or HER2-Fc once a week for four weeks. Trastuzumab (200 μg/inj) was used as a positive control of the experiment. Since the tumor size is proportional to the emitted luminescence, the efficiency of the different treatments was evaluated by measuring the luminescence once a week following i.p. injection of luciferin (100 μg/g). Mice were sacrificed when tumors reached the luminescence level of 3 × 107 photon/second. The results were expressed by an adapted Kaplan-Meier survival curve, using the time taken for a tumor to reach this level of emitted luminescence. Moreover, a median delay was defined as the time needed to 50% of the mice to reach that luminescence.
In vivoprophylactic vaccination with anti-Id scFv40 or anti-Id scFv69
MMTV.f.huHER2(Fo5) and FVB mice were immunized following the vaccination schedule described above when they were three months old. Follow-up of spontaneous development of mammary tumors was performed by direct palpation of the mammary glands for the entire duration of the experiment. Results were expressed by an adapted Kaplan-Meier curve of tumor-free survival using the age of palpable tumor onset in each mice. Moreover, the median delay was defined as the age when 50% of the mice had developed palpable mammary tumors. The study was ended at 59 weeks of age.
In vitrostimulation of splenocytes and analysis of IL2 and IFNγ production in culture supernatants
Six days after the final boost, the spleen of one immunized MMTV.f.huHER2(Fo5) mouse from each group was excised and pressed through stainless mesh. Isolated cells were washed with RPMI medium and then erythrocytes were lysed by hypotonic shock with 0.83% ammonium chloride solution. Splenocytes (5 × 105 per well) were seeded in flat-bottomed 96-well culture plates in triplicates in the presence of scFv69 and HER2-Fc (2.0 μg/ml/each) as stimulating agents. Mfe23-Fc fusion protein and scFv13R4 (2.0 μg/ml/each) were used as irrelevant proteins and 7.5 μg/ml Concanavalin A was used as a positive control. Cell-free supernatants were recovered 24 h and 72 h post-stimulation and were analyzed for IL2 and IFN-γ production by using the ELISA kit (BD Biosciences, San José, CA, USA) according to the manufacturer's instructions.
Immunoglobulin isotyping
Isotype distribution of anti-scFv69- and -HER2-Fc specific antibodies generated in immunized MMTV.f.huHER2(Fo5) and FVB mice was determined using a mouse monoclonal isotyping kit (Pierce, Rockford, IL, USA) according to the manufacturer's instructions.
Statistical methods
Concerning the in vitro assays, comparison of the percentage of cell viability inhibition between the different groups and normal mouse sera (NoMS) was performed with the Wilcoxon two-sample exact test for all continuous variables. The HER2-specific humoral responses in mice immunized with PBS or with scFv69 were compared with the Wilcoxon two-sample exact test. Differences were considered statistically significant when P < 0.05. Survival rates were estimated from the time of the xenograft until the date of the event of interest using the Kaplan-Meier method. Median survival was presented with 95% confidence intervals. The event of interest was a bioluminescence value of 3 × 107 photon/second for athymic nude mice and the appearance of palpable mammary tumors for huHER2-transgenic mice. Survival curves were compared using the Log-rank test. Statistical analysis was performed using the STATA 10.0 software (StataCorp LP, College Station, TX, USA).
Discussion
Several immunotherapeutic approaches targeting HER2 have been reported in the literature, including both passive and active immunization [
23]. The major problem in the use of HER2 as a target for active immunotherapy is the presence of immune tolerance to the self-antigen. The present study was conducted to assess the efficacy of two previously selected anti-Id (Ab2) human scFv fragments [
21] to generate an active anti-HER2 immune response. We show that vaccination with anti-Id scFv40 and scFv69 might represent a potential new therapeutic approach for the treatment of patients with HER2-positive tumors. We also demonstrate that the
in vivo anti-tumor effects of the anti-Id scFv69 vaccine are associated with a robust anti-HER2 humoral response through a Th2-dependent mechanism.
First, we demonstrate that sera from scFv40- or scFv69-immunized BALB/c mice contain Ab1' antibodies that strongly inhibit growth of SK-OV-3 cells, suggesting a "trastuzumab-like" biological effect as reported with the
in vitro use of the humanized anti-HER2 trastuzumab (Herceptin
®, ROCHE, Basel, Switzerland) [
24]. When adoptively transferred in nude mice bearing SK-OV-3 Luc tumor cells, these sera efficiently inhibited tumor growth. These results provide proof of the therapeutic potential of Ab1' antibodies. They also demonstrate the implication of the humoral response in the biological effects.
Second, we show that prophylactic vaccination of both virgin or primiparous MMTV.f.huHER2(Fo5) females with anti-Id scFv69 induces anti-HER2 Ab1' immune response followed by inhibition of spontaneous development of palpable tumors. The study by Finkle and colleagues indicated that early treatment with mu4D5 (the murine version of trastuzumab) was of benefit in MMTV.f.huHER2(Fo5)-transgenic females at high risk of developing huHER2-positive breast tumors. Similarly, we show that mice vaccinated with scFv69 (end of treatment at six months of age) were still free of tumors at the age of 14 months and that only one out of seven mice immunized with anti-Id scFv69 developed a tumor. Whereas in the work by Finkle et al. an early and prolonged treatment with mu4D5 was necessary to significantly alter mammary tumor incidence and progression, in our study, four injections of scFv69 were sufficient to elicit almost 100% protection against mammary tumors. Thus, different from the passive approach (mu4D5), active immunization with a therapeutic cancer vaccine may result in a gradual and lasting response, which could abolish tumor development and induce a long-term memory response leading to protection from disease recurrence.
Third, the analysis of the humoral response induced in immunized MMTV.f.huHER2(Fo5) mice shows that both a robust anti-scFv69 (Ab3) Th1 and an anti-HER2 (Ab1') Th2 response were associated with the delay of mammary tumor onset observed in animals treated with anti-Id scFv69. Based on these results, we think that anti-Id scFv69 could efficiently be used for the long term prevention of HER2-positive tumors via its specific anti-HER2 Th2-dependent immune mechanism.
ScFv40 and scFv69 were isolated by phage display, which is a relatively new technique to identify peptides or antibodies that mimic natural epitopes, including conformational B cell epitopes [
25], as it is the case for the epitope recognized by trastuzumab [
26]. The sequence analysis of scFv40 and scFv69 revealed no strong homology between their CDR1, 2 and 3 regions and the regions forming the epitope recognized by trastuzumab. This finding was expected as the epitope targeted by trastuzumab is described to be discontinuous. However, regarding the antigen mimicry capacity of the scFv69 fragment, its surface characteristics should be equivalent to those of the epitope of the selecting Ab1, although their amino acid sequence may differ [
27]. Similarly, the epitope mimics generated by Riemer
et al. [
28] by phage display using trastuzumab bear no sequence homology to HER2, but they are effective in mimicking the HER2 antigen. Their sequence was subsequently matched to the third loop of HER2 at the HER2/trastuzumab interface using computational methods [
29].
The low sequence homology between the CDR regions of scFv40 or scFv69 and the regions forming the HER2-epitope recognized by trastuzumab [
26] could explain the lack of anti-HER2 Th1 response following immunization with scFv69. Indeed, anti-Id scFv are supposed to mimic the three-dimensional rather than the primary sequence of HER2 which is in accordance with their anti-idiotypic nature. Furthermore, the lack of anti-HER2 Th1 response may be beneficial since it might balance the immunosuppressive effect of tumor-specific regulatory T cells (Treg). Depletion or blockade of Treg cells can lead to immune protection from tumor-associated antigens that are expressed as self-antigens [
30]. In our study, we demonstrate that scFv69-immunization leads to a HER2-specific Th2 immune response, suggesting that this type of vaccination could be effective in cancer patients even in the presence of immunosuppressive Treg cells.
Finally, cardiomyopathy has been shown to be a side effect of trastuzumab treatment in 7% of women following treatment with first-line anthracycline therapy and in up to 28% of women when trastuzumab is used concurrently with anthracycline [
31,
32]. In the study by Finkle
et al., human HER2 was reported to be expressed in heart, but no histological or clinical evidence of cardiac or any other organ toxicity were observed following treatment with mu4D5. However, these mice were not treated with other therapeutic agents, such as anthracyclines, before or during the study. In this context, it will be useful to study the effect of immunization with scFv69 on heart function, morphology and histology after treatment with anthracycline and trastuzumab as we believe that the most promising clinical application for anti-Id scFv69-based-vaccines should be as adjuvant therapy after treatment with chemotherapy and trastuzumab. Indeed, integrating a targeted vaccine therapy after induction of a major response with a monoclonal antibody (for example, trastuzumab) could be considered as another therapeutic approach [
33].
This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
MZL and TC made substantial contributions to the acquisition of data and to analysis and interpretation of data and drafting of the manuscript. SC made contributions to acquisition of data for the experiment of adoptive transfer of immune sera from anti-Id scFv-vaccinated mice. VG and BR made substantial contributions to conception, design and realization of in vivo experiments. CBM performed statistical analysis. IAA made substantial contributions to transgenic mice breeding and to in vivo vaccination in those mice. JPP and AP have given final approval of the manuscript version to be published. INT contributed to conception and design of the study, analysis and interpretation of data, and drafting and revision of the manuscript. All authors read and approved the final manuscript.