In pancreatic cancer, chemotherapy increases antitumor responses to tumor-associated antigens and potentiates DNA vaccination

Abstract

Background: Pancreatic ductal adenocarcinoma (PDA) is an almost incurable tumor that is mostly resistant to chemotherapy (CT). Adaptive immune responses to tumor-associated antigens (TAA) have been reported, but immunotherapy (IT) clinical trials have not yet achieved any significant increase in survival, confirming the suppressive environment of PDA. As CT has immune-modulating properties, we investigated the effect of gemcitabine (GEM) in antitumor effector responses to TAA in patients with PDA.

Methods: The IgG antibody repertoire in patients with PDA before and after CT was profiled by serological proteome analysis and ELISA and their ability to activate complement-dependent cytotoxicity (CDC) was measured. Peripheral T cells were stimulated in vitro with recombinant TAA, and specific proliferation, IFN-γ/IL-10 and CD8⁺/Treg ratios were measured. Mice that spontaneously developed PDA were treated with GEM and inoculated with an ENO1 (α-Enolase) DNA vaccine. In some experimental groups, the effect of depleting CD4, CD8 and B cells by specific antibodies was also evaluated.

Results: CT increased the number of TAA recognized by IgG and their ability to activate CDC. Evaluation of the IFN-γ/IL-10 ratio and CD8+/Treg ratios revealed that CT treatment shifted T cell responses to ENO1, G3P (glyceraldheyde-3-phosphate dehydrogenase), K2C8 (keratin, type II cytoskeletal 8) and FUBP1 (far upstream binding protein 1), four of the most recognized TAA, from regulatory to effector. In PDA mice models, treatment with GEM prior to ENO1 DNA vaccination unleashed CD4 antitumor activity and strongly impaired tumor progression compared with mice that were vaccinated or GEM-treated alone.

Conclusions: Overall, these data indicate that, in PDA, CT enhances immune responses to TAA and renders them suitable targets for IT.

Keywords: antigens; combination; drug therapy; immunotherapy; vaccination.

Figure 1

Analysis of the reactivity of sera from patients with PDA to recognize the PDA cell proteome and of their ability to induce CDC before and after CT. (A) Recognition of the proteome 2-DE map of the CFPAC-1 cell line by a representative PDA serum before and after CT. (B) Quantitative evaluation of increased IgG autoantibody recognition before and after CT. Results represent the mean of the percentage of all recognized spots. (C) Bar graph of LDH release measured as OD by damaged CFPAC-1 (black bars) and CAPAN-2 (white bars) cells subjected to CDC by patient sera before and after CT. All graphs indicate mean±SEM from 28 patients with PDA analyzed, and statistical significance is shown. CDC, complement-dependent cytotoxicity; CT, chemotherapy; LDH, lactate dehydrogenase; OD, optical density; PDA, pancreatic ductal adenocarcinoma; 2-DE, two-dimensional electrophoresis.

Figure 2

Analysis of the autoantibody response of patients with PDA after CT. Heatmaps representing the variation of antigen recognition level by autoantibodies measured by SERPA in sera from 28 patients with PDA receiving one round (left) or two rounds (right) of CT compared with before CT. Columns represent patients, and rows represent the antigens identified-listed in online supplemental table1A (all the isoforms of ENO1, FUBP1, K2C8 and G3P are indicated in red). Clustering analysis was performed considering data from patients receiving one CT round only (left). Rows and columns of the heat map reported in the right panel are sorted according to this clustering. CT, chemotherapy; PDA, pancreatic ductal adenocarcinoma.

Figure 3

Detection of autoantibodies to TAA before and after CT, and correlation with survival of patients with PDA. (A, B) Survival curves of patients with PDA (n=28) with increased (dotted line) or unchanged/decreased (solid line) of 2-DE western blot reactivity of autoantibodies to G3P_20 (A) and K2C8_16 (B) isoforms after CT. (C) ELISA detection of autoantibodies to ENO1, FUBP1, K2C8 and G3P in healthy subjects and sera of patients with PDA before and after CT. Each graph indicates the OD mean (n=28), and statistical significance is shown. CT, chemotherapy; OD, optical density; PDA, pancreatic ductal adenocarcinoma; TAA, tumor-associated antigens; 2-DE, two-dimensional electrophoresis.

Figure 4

Effect of combination of GEM and ENO1 vaccination in KC mice. (A) Schematic representation of mice treatment schedule. (B) Evaluation of the mean of tumor lesion diameter in treated mice sacrificed at 24 weeks of age. (C) ELISA detection of anti-ENO1 IgG titer (referred to as OD) in sera of mice throughout the treatment period. The experimental groups are indicated as follows: untreated (NT, white bars), GEM (light gray bars), ENO1 (dark gray) and GEM+ENO1 (black bars). (D) ELISA detection of anti-G3P IgG antibody (referred to as OD) in mice at 16 weeks of age. (E, F) ELISpot analysis of IFN-γ-secreting cells (indicated as number of specific spots) in the different treated groups at sacrifice after restimulation with ENO1 (E) or G3P (F) recombinant protein. (G–I) Immunohistochemical staining of CD4 (G), CD8 (H) and macrophages (I) in tumor lesions of mice at sacrifice. (L) Effect of the depletion of CD4, CD8 and B subsets in GEM+ENO1 or untreated mice, evaluated as the mean of tumor lesion diameter. (M) The spaghetti plot indicates the effect of combined treatment (GEM, green arrows; ENO1 vaccination, red arrows) on tumor growth, measured as tumor diameter. Treatment started when the mass of K8484 tumor cells injected subcutaneously reached 0.2 cm in diameter. The bar graph indicated the days required by the tumor mass to reach 1.0 cm of diameter in control and GEM+ENO1 experimental groups. In all experiments, the mice number per group was between 5 and 11; graphs report the mean±SEM values and statistical significance is shown. GEM, gemcitabine; OD, optical density.