Cellular antitumor immune response to a branched lysine multiple antigenic peptide containing epitopes of a common tumor-specific antigen in a rat glioma model

Abstract

Human malignant gliomas contain epidermal growth factor receptor (EGFR) gene mutations that encode tumor-associated antigens (TAAs) that can be targeted using immunological techniques. One EGFR mutant gene (EGFRvIII) encodes a protein with an epitope that is not found in normal tissues. A number of studies have focused on this unique epitope as a potential target for tumor vaccines. In the present study, we examined the cellular immune effects of a peptide containing multiple copies of the unique EGFRvIII epitope linked together by way of a lysine bridge. Fischer rats were vaccinated with an EGFRvIII multiple antigenic peptide (MAP). While vaccination produced a humoral immune response, anti-MAP antibody production was not accompanied by expression of the Th2 response cytokine IL-4. In MAP/GM-CSF vaccinated animals, a cellular immune response was detected in association with the appearance of CD4+ and CD8+ T cells at the tumor site. Splenocytes and CD8+ T cells from vaccinated rats produced the Th1 cytokine IFN-gamma in vitro in response to stimulation by rat glioma cells expressing EGFRvIII, but not by those expressing wild-type EGFR. MAP vaccine also induced a specific lytic antitumor CTL immune response against F98 glioma cells expressing EGFRvIII, but not against F98 cells expressing either wild-type EGFR or no receptor. The in vivo growth of F98(EGFRvIII) cells was attenuated in vaccinated rats; whereas, growth of F98(EGFR) cells was not. The median survival of vaccinated rats was increased 72% over that of unvaccinated controls challenged with intracerebral F98(EGFRvIII) tumor implants. Therefore, MAP vaccination produced a predominantly cellular antitumor immune response directed against F98 gliomas expressing the EGFRvIII target antigen. The potent immunosuppressive effects of F98 glioma cells mimic the human disease and make this particular tumor model useful for studying immunotherapeutic approaches to malignant gliomas.

Fig. 1

a Epitopes of the EGFRvIII fusion junction that are predicted to bind RT1.A¹. By virtue of an additional lysine (K) residue at the C-terminus of each monomer peptide, the MAP could present an altered peptide ligand capable of strong binding to RT1.A¹ and the CTL antigen receptor. Shaded boxes highlight the unique glycine residue (left) of the EGFRvIII fusion junction and the added lysine residue (right) in the MAP. b Amino acid sequence of the monomeric EGFRvIII peptide with glycine residue at fusion junction. c The EGFRvIII MAP contains an octavalent lysine backbone linked to eight peptide arms

Fig. 2

Western blot of proteins from rat F98 glioma cell lines transfected with expression vectors containing wild-type and EGFRvIII cDNA. Untransfected A431 human cervical carcinoma cells (lane 1) express the wild-type EGFR protein. F98_neo cells (lane 2) are transfected with vector alone, F98_EGFRvIII cells express human glioblastoma-derived 140-kDa EGFRvIII (lane 3), and F98_EGFR cells express the human wild-type EGFR (lane 4)

Fig. 3

Immunohistochemistry of cultured F98_EGFRvIII and F98_EGFR cells using antibodies to the ICD of EGFR and the ECD of EGFR as well as to FasL and TGFβ2. Antibody binding was visualized with DAB staining. Antibodies to murine IgG are shown as a negative control. Cells were imaged at ×400 magnification

Fig. 4

In vitro growth of individual clones of F98_neo, F98_EGFR, and F98_EGFRvIII cells

Fig. 5

a MAP binding activity in the serum of Fischer rats vaccinated with EGFRvIII MAP with or without GM-CSF as adjuvant. Vaccinated rats were without tumor cell implantation, except as indicated. A direct ELISA for MAP was used with rat anti-IgG antibody for detection. Serial serum dilutions were analyzed for anti-EGFRvIII antibodies and a 1:200 serum dilution is shown here. b IL-4 expression measured by ELISA in splenocytes in response to stimulation with peptide and F98_EGFRvIII cells

Fig. 6

Splenocyte proliferation was detected by Alamar blue reduction in culture medium. Splenocytes (1×10⁵) were obtained from rats following the indicated vaccinations and were cultured in the presence of MAP (10 μg) in DMEM with 10% FCS and IL-2 (10 U/ml) for the times indicated. Alamar blue reduction was quantified by measuring optical absorbance at 570 nm. Values are the mean of triplicate wells ± SE

Fig. 7a,b

Chromium release from F98_neo (a) and F98_EGFRvIII (b) glioma cells incubated at the indicated ratios with CTLs from Fischer rats vaccinated as indicated. ⁵¹Cr release was measured in culture supernatants by scintillation counting and revealed specific lysis of F98_EGFRvIII cells

Fig. 8

a IFN-γ production by splenocytes in response to stimulation by F98_EGFRvIII cells and EGFRvIII peptides. Splenocytes from MAP/GM-CSF–vaccinated rats were cultured in the presence of the indicated antigens for 4 days. After antigen stimulation, IFN-γ secretion was measured in ELISpot assays. The value for ConA is indicated at right. Values represent the mean of triplicate wells ± SE. b IFN-γ production by isolated CD4⁺ and CD8⁺ T cells from rats vaccinated with MAP/GM-CSF. Splenocytes from vaccinated rats were cultured in the presence of EGFRvIII MAP (10 μg) for 4 days. After specific antigen-induced expansion, CD4⁺ and CD8⁺ T cells cells were isolated and re-plated with the indicated secondary stimuli for 4 days. Values represent the mean of triplicate wells ± SE

Fig. 9

Immunofluorescence of CD4⁺ and CD8⁺ T cells in F98_EGFRvIII brain tumors of unvaccinated and MAP/GMCSF-vaccinated Fischer rats. Individually, neither MAP nor GM-CSF produced significant CD4⁺ or CD8⁺ T-cell infiltrates (data not shown)

Fig. 10

Subcutaneous F98 tumor size in vaccinated Fischer rats. After vaccination, Fischer rats were challenged with subcutaneous injections (1×10⁶ cells) of F98_neo, F98_EGFR, and F98_EGFRvIII clones. Tumors were measured on day 14 postimplantation, and volumes were calculated as described. Volumes reflect the mean of at least three tumors ± SE (p<0.006).

Fig. 11

a Growth curves of subcutaneous F98_EGFRvIII tumors in Fischer rats following immunization with MAP vaccine, with or without GM-CSF. After vaccination, Fischer rats were challenged with subcutaneous injections (1×10⁶ cells) of F98_EGFRvIII cells to assess the antitumor effect of MAP vaccination. Tumors were measured, and volumes were calculated as described. Values reflect the mean volume of at least three different tumors ± SE (p<0.006). b Volumes of F98_EGFRvIII tumors recorded 28 days after implantation with first vaccination given 4 days after implantation

Fig. 12

Kaplan-Meier survival curves of Fischer rats implanted intracranially with F98_EGFRvIII cells following immunization with MAP vaccine (No vaccine, n=10; GM-CSF alone, n=7; MAP alone, n=7; MAP + GM-CSF, n=8; p<0.001)