The epidermal growth factor receptor (EGFR) belongs to the erbB family of 4 closely related cell membrane receptors, also known as the Type I receptor tyrosine kinase family: EGFR or HER1/erbB1, first to be molecularly cloned,1 HER2/erbB2, HER3/erbB3 and HER4/erbB4. The 4 receptors consist of an extracellular ligand-binding domain (ECD), a transmembrane domain and an intracellular domain with tyrosine kinase activity for signal transduction. EGFR plays a central role in regulating both development and neoplastic processes. Binding of their specific ligands, such as epidermal growth factor (EGF) or transforming growth factor alpha (TGF-α) among others, induces receptor activation, modulation of cell proliferation and differentiation in normal tissues and tumors. Although expressed in nonmalignant cells, the EGFR can be found overexpressed or mutated in many human epithelial tumors such as breast,2, 3 lung,4 prostate5, 6 head and neck,7 colorectal,8 pancreatic,9 bladder,10 vulva and ovarian tumors.11 This overexpression has been correlated with disease progression and poor prognosis.12, 13 Activation of the EGFR signaling pathway in cancer cells has been shown to enhance cell proliferation, angiogenesis, tumor promotion and metastasis, and to decrease apoptosis. The potential of EGFR-targeted therapies for cancer treatment has increased the development of different passive agents. Passive immunotherapy with specific monoclonal antibodies (MAb)14, 15 and treatment with tyrosine kinase inhibitor drugs such as Iressa16, 17 and Tarceva18 are currently undergoing clinical trials with promising results or are commercially available. On the other hand, active immunotherapy strategies to block the EGF from binding to its receptor are being clinically tested by vaccinating patients with EGF coupled to P64k recombinant protein from Neisseria meningitidis.19
In addition, EGFR-based active specific immunotherapy may be an alternative and complementary approach for the treatment of epithelial tumors, provided that induction of an immune response against self EGFR is feasible. Preclinical studies of both a DNA vaccine based on xenogenic EGFR-ECD and dendritic cells pulsed with self EGFR-ECD have been recently published, demonstrating the validity of this active immunotherapy.20, 21 Here an alternative, more simple, approach, based on vaccination with the mEGFR-ECD protein for exploring the possibility of circumventing tolerance to self EGFR, was proposed. We constructed DNA plasmids encoding the murine EGFR-ECD, which were stably transfected in mammalian cells, and the corresponding recombinant protein was used for vaccination protocols. Besides, we expressed the Her1-ECD to compare the relative efficacy of self and non-self-immunization and for evaluating the immune response specificity to EGFR+ human tumor cells. A strong DTH response and specific IgG titers with a TH1-associated subclass pattern were obtained by inoculation of mice with the self protein in very small sized proteoliposomes (VSSP) and complete Freund adjuvant. The corresponding immune sera showed in vitro antitumor effect, inhibiting EGFR+ tumor cells proliferation. mEGFR-ECD vaccination induced a potent antimetastatic effect in 3LL-D122 Lewis lung carcinoma. Antibodies obtained in mice immunized with Her1-ECD/VSSP evidenced a low cross-reaction with the parent mEGFR-ECD.
Material and methods
Construction of the expression vector encoding mEGFR-ECD and Her1-ECD
DNA encoding the extracellular domain of murine EGFR was amplified by PCR using total cDNA from mouse liver as template. The sense primer 5′-CGGAATTCCTCTCCCGGTCAGAGATGCGAC-3′ includes EcoRI excision site, the initiation codon and 4 bp from EGFR signal sequence. The antisense primer 5′-CGGGATCCTCAAGATGGTATCTTTGGCCCAGATG-3′ is complementary to bp 1978–2000 in 3′ region and contains a stop codon (double underlined) and a BamHI excision site (single underlined). The PCR product, a 1.9-kb fragment, was cloned into EcoRI/BamHI sites of the pBluescript KS+ vector. The fragment encoding for mEGFR-ECD was recovered using HindIII/BamHI enzymes and cloned into the pcDNA3 expression vector (Invitrogene, San Diego, CA), generating the mEGFR-ECD/pcDNA3 plasmid.
DNA encoding the extracellular domain of human EGFR (Her1-ECD) was amplified by PCR using the Her1Δ533/pRK5 plasmid as template. The sense primer 5′-GGGGTACCCTTCGGGGAGCAGCGATGCGA-3′ includes a KpnI excision site (underlined), the initiation codon ATG and 3 bp from the signal sequence of Her1. The antisense primer 5′-GCTCTAGATCAGGACGGGATCTTAGGCCCA-3′ is complementary to bp 2103–2123 in the3′ region, and contains a stop codon (double underlined) and an XbaI excision site (single underlined). The PCR product, a 1.9-kb fragment, was cloned into KpnI/XbaI sites of the pcDNA 3-expression vector, generating the Her1-ECD/pcDNA3 plasmid.
mEGFR-ECD and Her1-ECD sequences were confirmed, by dideoxy nucleotide sequencing analysis, to be identical with those previously reported.22, 23 All enzymes were supplied by Boehringer-Mannheim, Penzberg, Germany.
Ehrlich Ascites tumor (EAT, ECACC No. 87032503), 3LLD122, a metastatic variant of Lewis lung carcinoma,24 the murine thymoma EL4 (ATCC TIB-39), human embryonic kidney (HEK293, ATCC CRL-1573), human epidermoid carcinoma A431 (ATCC CRL-1555) and human lung adenocarcinoma H12525 cell lines were grown in DMEM (Gibco, NY, USA) supplemented with 10% fetal calf serum (FCS) (Hyclone, Utah), 2 mM L-glutamine, 1 mM sodium pyruvate, penicillin 100 U/ml and streptomycin 100 μg/ml (Life Technologies, Grand Island, NY). HEK293 transfectants were adapted to growth in HyQ PF 293 (protein-free medium from Hyclone, Utah).
Generation of HEK293 transfectants
HEK293 cells were grown in 6-well plates (1.75 × 105 cells/ml), and 8 hr later were transfected with 4 μg of mEGFR-ECD/pcDNA3 or Her1-ECD/pcDNA3 plasmids, using the calcium phosphate transfection system. Plates were incubated overnight at 3% CO2, and then at 5% CO2. Transfected cells were selected in medium containing 1,000 μg/ml of G418 (Geneticin, Sigma) starting 48 hr after transfection, for the generation of mEGFR-ECD/HEK293 and Her1-ECD/HEK293 stable cell lines. Mock transfection (with pcDNA3 vector) was used as a negative control.
Lectin- or antibody-mediated precipitations of therecombinant proteins
Supernatants from mEGFR-ECD/HEK293 or Her1-ECD/HEK293 cultures (2 ml) were mixed with 10 μl of lectin-agarose (a lectin from Triticum vulgaris, Sigma, St. Louis) or 1 μg of R3 MAb (a MAb specific for human EGFR extracellular domain; CIM, Havana, Cuba) plus 20 μl of Protein A-Sepharose (Amershan-Pharmacia Biotech, Uppsala, Sweden), respectively. Samples were gently shaken overnight at 4°C, and afterwards centrifuged for 1 min at 11,000g. The precipitated recombinant proteins were separated on SDS-PAGE 7.5% and visualized by silver staining.
Purification and immunoblotting of mEGFR-ECD and Her1-ECD
Recombinant proteins were purified from confluent cultures of the respective transfectants by affinity chromatography. EAH-Sepharose 4B (Amershan-Pharmacia Biotech) was covalently coupled to human recombinant EGF (hrEGF) (Center of Genetic Engineering and Biotechnology, CIGB, Cuba) or to R3 MAb for mEGFR-ECD and Her1-ECD purification, respectively. Equilibration and washing steps were performed with PBS/NaCl (1 M, pH 7.0) and protein elution with glycine (0.2 M, pH 2.8). Purity was assessed by densitometry, using a personal densitometer SI (Amershan-Pharmacia Biotech) and Imag Quant Software. Protein concentrations were assayed by Lowry's method.26
Purified proteins' identity was established by immunoblotting. Recombinant proteins (30 μg) were applied to 7.5% SDS-PAGE gels and transferred to PVDF membranes (Gelman, Ann Arbor, MI). After blocking with NEGT buffer (0.15 M NaCl, 5 mM EDTA, 500 mM Tris-HCl (pH 7.5), 0.02% Tween 20, 0.04% gelatin), membranes were incubated with R3 (for Her1-ECD, data not shown) or 7A7 (for mEGFR-ECD) MAbs and proteins were visualized using horseradish peroxides-conjugated secondary antibodies followed by enhanced chemiluminescence (Perkin Elmer Life Sciences).
Total RNA was isolated from Balb/c or C57BL/6 mice thymus using Trizol reagent (Life Technologies) according to the manufacturer's instructions. The reverse transcription and polymerase chain reaction (RT-PCR) was performed using the SUPERSCRIPT™ one-step RT-PCR system. Used EGFR and β-actin primers were designed from the published sequences.27 After PCR amplification, 10 μl of the RT-PCR products were separated by electrophoresis on 1.5% agarose gels and visualized with ethidium bromide. Total RNAs from murine thymoma EL4 was used as negative control.
Mice and immunization protocols
Female C57BL/6 mice, aged 8–12 weeks, were purchased from the National Center for Laboratory Animals Production (CENPALAB, Havana, Cuba). All mice were kept under pathogen-free conditions. Animal experiments were approved by the Center of Molecular Immunology's Institutional Animal Care and Use Committee (CIM).
Mice (n = 5 or n = 10 for DTH and humoral response studies, respectively) were immunized with 50 μg of either mEGFR-ECD or Her1-ECD in FA adjuvant, complete for the first immunization and incomplete for the rest (Sigma) or in VSSP adjuvant, obtained from the combination of the outer membrane proteins of Neisseria meningitidis with GM3 ganglioside, in water/oil (Montanide ISA 51, Seppic, Paris, France) emulsion.28 Immunizations were made, subcutaneously (sc) for FA-adjuvated or intramuscularly (im) for VSSP-adjuvated preparations, on days 0, 14 (for the DTH study) and 0, 14, 28 and 42 (for humoral response studies). Sera were extracted on days 0, 21, 35 and 56. Control groups received PBS/FA or PBS/VSSP.
Seven days after the last immunization, mice were sensitized by intradermal injection with 50 μg of mEGFR-ECD in 50 μl of PBS in the right hind foot pad and by the same volume of PBS in the left foot pad. After 48 hr, mice foot swellings were measured using a plethysmometer (Ugo Basile, VA, Italy). Mice immunized with 100 μg of Keyhole Limphet Hemocyanin (KLH), (Sigma, Aldrich) in FA and sensitized with KLH in PBS were used as positive controls, while those injected with PBS/FA and sensitized with mEGFR-ECD were considered as negative controls. Differences in DTH between treatment groups were statistically validated by Kruskal Wallis and Dunn's multiple comparison test.
Microtiter plates (High binding, Costar) were coated with 10 μg/ml of mEGFR-ECD or Her1-ECD in carbonate buffer (0.1 M, pH 9.6) and incubated overnight at 4°C. Plates were blocked with 5% calf serum in PBS/Tween-20, and sera dilutions in duplicate, from immunized mice (n = 10), or preimmune sera (as negative control) were incubated for 1 hr at 37°C. Alkaline phosphatase-conjugated goat anti-mouse IgG antibody (Sigma) was added and incubated for 1 hr at 37°C. After addition of p-nitrophenylphosphate (1 mg/ml) (Sigma), the optical density (OD) was measured at 405 nm using a microwell system reader (Organon Teknica, Salzburg, Austria). All washes were made with PBS/Tween-20. The Mann Whitney U test was used to assess statistical differences between individual time points in the humoral response kinetic. ELISA test background was 2 times the OD at 405 nm of preimmune sera, which coincides with the OD value for PBS.
For determination of serum IgG subclasses, secondary isotype-specific biotinylated rat anti-mouse IgG1, IgG2a, IgG2b or IgG3 antibodies were used (PharMingen). Optimal secondary reagent dilutions were established by ELISA with 14F7 (IgG1-specific for NGcGM3), T3 (IgG2a-specific for CD3) and T4 (IgG2b-specific for CD4) MAbs (CIM), while R24 MAb (IgG3-specific for GD3) (kindly provided by Dr. Philip O. Livingston, Memorial Sloan Kettering Cancer Center, NY). Unpaired t-test was used to check statistically significant differences between sera dilutions.
Specific inguinal lymph (LN) nodes spot forming cells (SFC) were obtained from mice immunized with Her1-ECD/VSSP and tested by enzyme-linked immunospot (ELISPOT) as previously described,29 with some modifications. Maxisorp 96-well plates (Nunc) were coated with 10 μg/ml of Her1-ECD or mEGFR-ECD in 50 μl carbonate buffer (pH 9.8) at 4°C overnight. After blocking with 5% BSA in PBS, different dilutions of pooled LN cells were incubated in triplicate for 6 hr at 37°C in a 5% CO2 incubator. Antibodies secreted by individual cells were revealed as spots by the stepwise addition of 1.5 μg/ml of alkaline phosphatase-conjugated goat anti-mouse IgG (Fcγ) or IgM (Fcμ) antibodies (Jackson Immunoresearch laboratories, West Grove, PA) and the addition of 1 mg/ml of 5-bromo-4-chloro-3-indolyl phosphate (BCIP) substrate (Sigma, San Louis, MO) in 0.1 M AMP buffer (pH 10.5) containing 0.6% agarose. Plates were incubated overnight at 4°C and the results were scored the next day by counting the number of specific SFC in a stereoscopic microscope. LN cells from mice injected with PBS/FA were used as negative control.
FACS analysis for EGFR recognition
Cells were stained with sera from immunized mice (1/200 dilution) followed by FITC-goat anti-mouse IgG (Jackson Immunoresearch laboratories). Up to 10,000 cells were acquired using a FACScan flow cytometer and analyzed using the CellQuest software (Beckton Dickinson, San Jose, CA). PCR and 7A7 MAb,30 which is specific for mEGFR-ECD, were used to confirm EGFR expression in murine cells. Human EGFR expression in the corresponding cells was confirmed with R3 MAb. EL4 murine cell line and human lymphocytes were used as negative control cells.
Flat-bottomed 96-well microculture plates were seeded with 104 cells in 100 μl/well and grown in DMEM supplemented with 1% FCS in the presence of sera dilutions. After 48 hr of incubation at 5% of CO2, cells' viability was measured by the modified colorimetric MTT (3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazodium bromide) assay.31, 32 Media were replaced by 100 μl/well of MTT (1 mg/ml) and plates were incubated under culture conditions for 4 hr. Formazan crystals were dissolved by addition of 100 μl/well of dimethyl sulfoxide followed by 30 min incubation at 37°C. Absorbance (OD) was measured at 540 nm using a microplate spectrophotometer and the reference wavelength (620 nm) OD subtracted. Background control contained only culture medium without cells. Cells without treatment were included as a maximum cell growth point. R3 and 7A7 MAbs were used as positive controls for human and murine cells lines, respectively. Statistical differences in the in vitro viability assay were evaluated by the unpaired t-test.
Flat-bottomed 12-well microculture plates were seeded with 5 × 105 cells in 100 μl/well and grown in DMEM supplemented with 1% FCS in the presence of diluted sera (1/10). Sera from nonimmunized mice were used as control for unspecific complement-mediated cytotoxicity. To measure specific complement-independent cytotoxicity, immune sera were heated for 30 min at 56°C, and after 24 hr of incubation at 5% of CO2, death cells were counted by FACS using propydium iodide. Cells without treatment were included as control for minimum death cell. R3 and 7A7 MAbs were used as positive controls for human and murine cells lines, respectively.
Tumor challenge assay
C57/BL6 mice (n = 10) were immunized, intramuscularly, 3 times biweekly with 100 μg of mEGFR-ECD/VSSP or PBS/VSSP. One day before the second immunization, mice were challenged with 2 × 105 tumor cells, subcutaneously, in the foot pad. Three weeks later, tumors reached 0.8 cm and were surgically removed. Twenty-one days after surgery, mice were sacrificed and spontaneous lung metastases quantified by weighing the lungs. Statistical differences between groups were determined by unpaired t-test.
Reproductive side effects studies
Female Balb/c mice (n = 10) were immunized with mEGFR-ECD/VSSP or PBS/VSSP (control group) as previously described. After checking the induction of specific antibodies against mEGFR-ECD, mice were mated with nonimmunized male animals. Fertility (number of mice completing pregnancy), number of pups, pups' birth weights and certain postnatal developmental features such as eyes opening, hair growth and incisor eruption were observed. The Mann Whitney U test was used to test statistically significant variations in the reproduction parameters between treated and control animals.
Variance homogeneity and data normal distribution were analyzed by Bartlett's and Bonferroni tests, respectively, using the SPSS version 10.0 software. All statistical tests (Kruskal Wallis and unpaired t-tests) were 2-sided and conducted using the Graph Pad Prism version 4.00 software. A probability value of p < 0.05 was considered as statistically significant.
mEGFR-ECD and Her1-ECD expression and purification
cDNAs encoding mEGFR-ECD and Her1-ECD were successfully cloned into the pcDNA3 expression vector (Fig. 1a) and transfected in HEK293 cells as previous description. Expression of the soluble recombinant proteins by stable HEK293 transfectants was checked by lectin-agarose precipitation (mEGFR-ECD) or immune-precipitation with R3 MAb (Her1-ECD). In each case, as expected, a 105-kDa protein band was displayed but not for mock transfection, as determined by SDS-PAGE. Figure 1b shows the corresponding results for the mEGFR-ECD. mEGFR-ECD identity was further confirmed by affinity chromatography (with hrEGF-EAH Sepharose) purification combined with SDS-PAGE and western blotting, showing again the 105-kDa protein band (Fig. 1c). Achieved protein purity was 98% after only 1 purification step as determined by densitometry (data not shown).
mEGFR is expressed in thymus
As has been previously reported from studies with rat cell lines33 and human thymus,34 the EGFR presence in mice thymus was demonstrated by RT-PCR. The corresponding EGFR cDNA band (Fig. 2) became apparent when thymuses from C57BL/6 and Balb/c mice were analyzed, but not from murine thymoma EL4, used as a negative control.
mEGFR-ECD immunization induces DTH response
Generation of specific DTH responses was considered as a primary endpoint for the mEGFR-ECD vaccine (emulsified in FA) ability in immunizing mice. Mice were subcutaneously injected with mEGFR-ECD/FA and then sensitized with 50 μg of the nominal antigen. After 48 hr, mice foot swellings were measured, and inflammation scores in the vaccinated group animals were higher than those of the negative control group (p < 0.05, Dunn's multiple comparison test). Noteworthy, DTH responses induced in mice vaccinated with either mEGFR-ECD or KLH, a strongly immunogenic protein, were similar (p > 0.05) (Fig. 3).
mEGFR-ECD immunization induces a strong and long-lasting specific humoral response
Vaccine-related humoral responses were explored by immunizing C57BL/6 mice 4 times biweekly with 50 μg of mEGFR-ECD in 2 different adjuvants: FA, the reference adjuvant, and VSSP, a new product already clinically tested in humans. Inoculated mice developed high serum IgG antibody levels against the immunizing protein, which increased with successive immunizations, for both adjuvant formulations (Fig. 4). Indeed, the mEGFR-ECD/VSSP vaccine induced higher antibody titers than the FA one in each sera collection day (Mann Whitney U test, p < 0.05). Eight of 10 mice (80%) immunized with mEGFR-ECD/VSSP showed specific antibody titers above 1/40,000 by day 56, even reaching values up to 1/160,000, whilst only 2 of 10 mice (20%) rose titers above1/40,000 in the mEGFR-ECD/FA immunized group (Table I). One year after having finished the immunization schedule and without antigen recall, the sera-specific IgG levels decreased in both groups of mice, although an appreciable response was still detected in 100% of animals (1/1,000 and 1/100 sera dilutions) (data not shown).
Table I. Response Frequency and IgG Titers1 in Individual Animals by Day 56 After Immunization with the mEGFR-ECD, Using FA or VSSP as Adjuvants
Assayed by ELISA.
mEGFR-ECD in FA
mEGFR-ECD in VSSP
Administration of mEGFR-ECD in FA or VSSP to Balb/c mice produced elevated IgG titers in 100% of animals, similarly to what was observed in the C57Bl/6 case (data not shown).
Vaccination with mEGFR-ECD polarizes systemic immunity to a TH1 pattern
Consistent with previous results demonstrating that complete FA preferentially promotes a TH1 type response to the accompanying antigen,35 elevated levels of IgG2a, IgG2b, and IgG1 were detected in day 21 sera corresponding to mice immunized with mEGFR-ECD/FA. As shown in Figure 5, while no differences in IgG2a levels were found in sera of mice vaccinated with FA or VSSP formulations (unpaired t test, p > 0.05), the use of VSSP promoted the induction of higher levels of IgG2b (p < 0.05).
Her1-ECD/VSSP immunization generates specific B cell clones with low cross-reactivity to the mEGFR-ECD
We wonder if immunization with the xenogenic EGFR-ECD could induce antibodies reacting with the self EGFR. C57BL/6 mice were immunized with 50 μg of Her1-ECD in FA or VSSP. All immunized mice, independent of the used adjuvant, developed high IgG antibody titers against the human protein (1/320,000) by day 56 (data not shown). Alternatively, ELISA experiments reflected an evident but low cross-reactivity with the murine protein by day 21 (Fig. 6a). This low cross-reactivity against the mEGFR-ECD was confirmed by ELISPOT assay. Sixty-two IgG and 6 IgM secreting specific SFC in 106 LN cells were found in LN of Her1-ECD-immunized mice, while only 21 IgG secreting SFC/106 LN cells cross-reacted with the mEGFR-ECD (Fig. 6b).
Immune sera recognizes full length EGFR by FACS
To check whether immunizations with a truncated EGFR affected the recognition of the full length EGFR in its native conformation on the cell surface, EGFR+ cells were analyzed by FACS. EAT36 and 3LL-D122 murine cell lines were positively stained by sera from mice immunized with mEGFR-ECD/VSSP (Fig. 7a). Besides, sera from mice immunized with Her1-ECD/VSSP readily reacted with A431 and H125 cell lines (Fig. 7b). Sera from control mice, immunized with PBS/VSSP, neither recognized murine nor human EGFR+ cells.
Immune sera inhibit EGFR+ tumor cells growth and possess cytotoxic effect
To determine whether immunization with mEGFR-ECD/VSSP or Her1-ECD/VSSP can generate serum antibodies affecting murine or human tumor cells growth in vitro, the MTT viability assay was performed. Incubation of 3LL-D122 or H125 cells with sera obtained from mice immunized with mEGFR-ECD/VSSP or Her1-ECD/VSSP, respectively, decreased the number of viable cells if compared with preimmune sera after 48 hr, and this effect was sera-dilution-dependent (unpaired t-test, p < 0.05) (Fig. 8). In addition, the immune sera in vitro cytotoxicity over EGFR+ cells was evaluated after treating cells for 24 hr with the corresponding complement inactivated samples, following propydium iodide staining. FACS analysis showed that treatment of 3LL-D122 cells with inactivated sera produced 55.83% of death cells, suggesting that a complement-independent cytotoxicity mechanism is operating (Fig. 9a). In parallel, H125 cells were treated with sera from mice immunized with Her1-ECD/VSSP, and the effects over the cells evaluated. Incubation with immune sera manifested cytotoxic effects (FACS) (Fig. 9b).
Antimetastatic effect of mEGFR-ECD/VSSP vaccination
To investigate whether the autologous vaccination can protect individuals from metastatic widespread, mice were immunized with the mEGFR-ECD in VSSP and, 1 day before the second immunization, challenged with 2 × 105 tumor cells in the foot pad. Three weeks after malignant cells inoculation, tumors were surgically removed. Twenty-one days after surgery, mice were sacrificed and the spontaneous lung metastases quantified by weighing the lungs. As shown in Figure 10, vaccination of mice significantly reduced lung metastasis (p < 0.01), compared with animals in the control group.
Absence of reproductive side effects in mice immunized with mEGFR-ECD/VSSP
The potential side effects of “self” immunization in humans were stressed by examining the appearance of possible toxic symptoms in female mice immunized with the mEGFR-ECD in VSSP. After the vaccine, inoculation animals were mated and their progeny studied. Pregnancies rates were 5 out of 10 in the immunized group, while 3 out of 10 in control mice. The median number of pups per litter in both groups was 6 (range 5–7). Features like newborns' weights, hair growth, eyes opening and incisor appearance did not comparatively differ (Mann Whitney U test, p > 0.05) (Table II). On the other hand, a group of mice were observed for 1 year after fulfilling the immunization protocol, and the vitality, temperature and food intake were completely normal, without changes in functional hepatic parameters when compared with those of nonimmunized mice (data not shown).
Table II. Pregnancy and Newborn Reproductive Parameters Measured After mEGFR-ECD/VSSP Immunization
Despite the EGFR wide expression in the organism, it can be considered as a tumor-associated antigen (TAA) because of its overexpression in many epithelial tumors,37 its implication in tumor growth and correlation with bad prognosis.12 For that reason, the EGFR has become an attractive target for cancer therapy, and many attempts are currently ongoing, using this molecule as target for passive therapy with tyrosine kinase inhibitors and MAb.38, 39, 40 Our results demonstrated that the murine EGFR extracellular domain, in an appropriate adjuvant, is enough immunogenic in mice to promote a strong antimetastatic effect in a relevant EGFR+ tumor model, highlighting this particular approach as an attractive new strategy for cancer treatment.
TAA tolerance represents a significant challenge for effective immunotherapy of human cancer, and in any successful vaccine strategy, this issue should be conveniently addressed in advance. The experimental demonstration that breaking self tolerance is possible has been recently published and is generally accepted under the principle that self recognition is a physiological phenomenon.41 The fact that the EGFR is widely distributed in normal epithelial tissues hardly suggests that this molecule could be a very poor immunogen, introducing additional difficulties to any vaccine design. In our vaccine approach, the construction of chimerical molecules was avoided, introducing instead potent TH1 adjuvants. Emulsifying the recombinant protein either in FA or VSSP (a product already clinically tested in humans) renders a rather simple formulation. As the primary endpoint for the vaccine induced mEGFR-ECD immunogenicity, the induction of DTH was considered. Surprisingly, mEGFR-ECD/FA vaccination promoted severe inflammations in mice foot pads after the specific sensitization, similar to those mice previously immunized with KLH/FA and sensitized with this “foreign” protein.
Immunization of mice with the mEGFR-ECD in FA or VSSP also stimulated the specific humoral immunity, characterized by elevated IgG antibody titers, successively incremented with reimmunizations, an indicative of a mature response. Both adjuvants influenced IgG subclasses distribution in favor of IgG2a and IgG2b, an indirect indication of TH1 differentiation. Particularly, higher levels of IgG2b were associated with VSSP formulations. An indicative that the immune reaction against this kind of self proteins is limited came from other member of the EGFR family, Her2. Monkeys were successfully immunized only after 6 immunizations with the Her2-ECD, formulated in the powerful adjuvant Detox, and specific IgG titers never reached 1/10,000,42 a value lower than that induced in mice with the mEGFR-ECD in FA or VSSP. Besides, Disis et al.43 have shown that a Her2 neu-peptide-based vaccine, but not a whole-protein vaccine, can elicit humoral and cellular responses in rats. The EGFR is a tolerated self-antigen for which a specific T cells thymic deletion mechanism could be operating, requiring the presence of the auto-antigen in the thymus.44, 45 The EGFR thymic expression, although earlier reported in humans34 and rats,33 was confirmed in our lab for Balb/c and C57Bl/6 mice by RT-PCR. Similar to what has been reported in normal people, a total absence of anti-EGFR natural auto-antibodies was also observed in mice sera by ELISA and FACS. Nevertheless, evidences coming from cancer patients indicated that the presence of discrete serum anti-EGFR natural antibodies can be detected with certain frequency.46 The same observations have been reported for breast cancer patients in which anti-Her2 antibodies and CTL could be measured,47 lacking enough efficacy in preventing tumor progression. This “natural” immunity to Her2, present only in a minority of patients overexpressing the receptor, is of low magnitude.48 The low, natural immune responses to Her1 and Her2 in cancer patients means that any associated target-directed therapeutic vaccine must efficaciously stimulate naïve B and T lymphocytes. In fact, Her2 vaccines, constructed with synthetic peptides mixed with granulocyte-macrophage colony stimulating factor as adjuvant, have been clinically tested in breast, ovarian and lung cancer patients overexpressing Her2 and 68% of them developed T-cell immunity against the self Her2 protein.49
More likely explanations for the unusually strong immunogenicity of the autologous Her1 protein observed in this study might be the full length EGFR truncation, the adjuvant conditioning of the antigen presentation context or both. EGFR truncation could modify the T cells repertoire immunodominance, favoring the presentation of cryptic determinants.50 In this case, protein truncation did not affect the full length EGFR recognition in its natural conformation in the cells by the vaccine-induced serum antibodies, as determined in FACS experiments. This result suggests certain differences with the Her2 model, in which specific antibodies were undetectable in sera obtained from rats immunized with the rat Her2 intracellular domain (ICD), while CTL and antibodies with degenerated specificities for the human and rat Her2/neu were produced when the inoculated immunogen was the highly homologous foreign human ICD.51
The use of potent TH1-type adjuvants in combination with poorly immunogenic self proteins to promote a proinflammatory context for loosing the regulatory cells circuit was an attractive, tested idea. In this sense and as usual in experimental vaccine approaches complete FA was selected as reference TH1 adjuvant.35 While complete FA cannot be used in human vaccines, a new, already clinically tested adjuvant (VSSP) with peculiar immune-modulatory properties was introduced in the EGFR-ECD vaccines. VSSP monotherapy in mice induced elevated IgG levels with a TH1-related pattern against GM3, a poorly immunogenic ganglioside,52 and dendritic cell maturation with IL-12 production,52 which in turn is pivotal for proinflammatory responses.53, 54
Although the amino acid sequence homology between human and murine EGFR-ECDs is about 87%, sera obtained from mice immunized with the Her1-ECD formulation were unable to appropriately react with the murine protein, while only 1/3 of the vaccine-stimulated B cells, secreting specific IgG antibodies, cross-reacted with the mEGFR-ECD. The finding that serum-specific antibodies, induced in mice by immunization with the corresponding EGFR-ECDs in VSSP, caused the slow growing of EGFR+ tumor cells and also promoted a strong target-directed complement-independent cytotoxic effect emphasize the quality of the vaccine-induced immune response.
Further results associated with the in vivo mEGFR-ECD/VSSP antimetastatic effect in the EGFR+ 3LL-D122 Lewis lung carcinoma model are encouraging if the intermediate character of the selected experimental setting (rather therapeutic than prophylactic) was considered. Even though in this case tumor cells were inoculated into mice foot pads just after vaccine priming, supplying afterwards the 2 booster injections, a significant decrease in lung metastases number, after primary tumor surgical removal, was evident.
One week after the third vaccine administration, and coincident with the surgical removal of primary tumors, specific IgG titers in mice sera were elevated as noticed from the humoral response kinetics, probably suggesting that the induced antibodies might have a role in avoiding tumor cells' dissemination or in turn the growth of malignant cells already lodged in the lungs, or both. Future studies stressing the relative contribution of humoral and cellular immunity in the vaccine-induced antimetastatic effect, through hyperimmune serum transfer and the appropriate lymphocyte subsets depletion experiments are currently ongoing. The “spontaneous metastasis” 3LL Lewis lung carcinoma model, employed in this study, is rather significant because of the most likely resemblance to the real clinical situation in which surgeons frequently remove patients' primary tumors successfully, but unfortunately and more commonly, disease spreading will follow. Indeed, avoiding distant metastatic dissemination could be the most appropriate task for effective cancer vaccines. Interestingly, other 2 active immunotherapy approaches targeting the EGFR have shown recently in vivo efficacy in the Lewis lung carcinoma model but in a primary tumor scenario. A DNA vaccine,20 based on the xenogenic EGFR-ECD gene and a self EGFR-ECD protein pulsed dendritic cells vaccine21 were able to keep alive 60% of mice preventatively injected and afterwards challenged with LL/2c tumor. Although together all these results strongly suggest that EGFR could be a significant target not only for passive but also for active immunotherapy, another crucial remaining question is the better vaccine approach to follow up in the upcoming future clinical trials. A relevant learning from the present work is that just the use of the recombinant self protein in a potent adjuvant, like VSSP, could be appropriate for establishing an antitumor immunity in patients with EGFR+ tumors, indicating the existence of a new opportunity in this particular target, different from sophisticated vaccine technologies like the dendritic cells approach, or up to now ineffective vaccine formulations in humans like naked DNA.
EGFR-targeted therapies are expected to produce side effects related to the induction of autoimmunity. In fact, some side effects as skin rash have been reported for some related drugs,55 but important toxic effects have not been found for the majority of the different approaches tested in clinical trials. As an example THERACIM, an anti-EGFR MAb (humanized R3, CIM), has been clinically tested (Phase II trials), in combination with radiotherapy, in head and neck cancer patients, raising up to 600 mg/cycle, without the detection of skin rash symptoms. Although it has been reported that EGFR gene expression inhibition is critical for cancerous cell growth but not for normal cells,56 active immunization, providing a long lasting specific immune response, should be carefully monitored for possible side effects. The conducted immunization experiments with mEGFR-ECD/VSSP showed that while 1 year after the last vaccine booster anti-mEGFR-ECD IgG low levels were detectable in most animals, signs of toxicity were absent and functional hepatic parameters behave as in naïve mice.
Induction of EGF deficiency in rats affects the development of fetal but not adult tissues.57 Considering the role of EGF in the epigenetic regulation of fetal and neonatal development, we studied the effect of anti-EGFR immunity in female mice fertility and their progeny. Marked side effects in these animals as a consequence of vaccination were not detected.
We conclude that self EGFR-ECD is rather immunogenic in FA or VSSP contexts. Taken together, the immunization approach described here may be an attractive and novel strategy for EGFR+ cancer active immunotherapy.
We thank Armando López (CIM, Cuba) for his technical assistance in animal's experiments, and also Dr. Donald Morton (John Wayne Cancer Institute, USA) and Dr. Agustín Lage (CIM, Cuba) for their critical revision of the manuscript and useful suggestions.