Optimizing parameters for clinical-scale production of high IL-12 secreting dendritic cells pulsed with oxidized whole tumor cell lysate

Abstract

Background: Dendritic cells (DCs) are the most potent antigen-presenting cell population for activating tumor-specific T cells. Due to the wide range of methods for generating DCs, there is no common protocol or defined set of criteria to validate the immunogenicity and function of DC vaccines.

Methods: Monocyte-derived DCs were generated during 4 days of culture with recombinant granulocyte-macrophage colony stimulating factor and interleukin-4, and pulsed with tumor lysate produced by hypochlorous acid oxidation of tumor cells. Different culture parameters for clinical-scale DC preparation were investigated, including: 1) culture media; 2) culture surface; 3) duration of activating DCs with lipopolysaccharide (LPS) and interferon (IFN)-gamma; 4) method of DC harvest; and 5) cryomedia and final DC product formulation.

Results: DCs cultured in CellGenix DC media containing 2% human AB serum expressed higher levels of maturation markers following lysate-loading and maturation compared to culturing with serum-free CellGenix DC media or AIM-V media, or 2% AB serum supplemented AIM-V media. Nunclon™Δ surface, but not Corning(®) tissue-culture treated surface and Corning(®) ultra-low attachment surface, were suitable for generating an optimal DC phenotype. Recombinant trypsin resulted in reduced major histocompatibility complex (MHC) Class I and II expression on mature lysate-loaded DCs, however presentation of MHC Class I peptides by DCs was not impaired and cell viability was higher compared to cell scraping. Preservation of DCs with an infusible cryomedia containing Plasma-Lyte A, dextrose, sodium chloride injection, human serum albumin, and DMSO yielded higher cell viability compared to using human AB serum containing 10% DMSO. Finally, activating DCs for 16 hours with LPS and IFN-γ stimulated robust mixed leukocyte reactions (MLRs), and high IL-12p70 production in vitro that continued for 24 hours after the cryopreserved DCs were thawed and replated in fresh media.

Conclusions: This study examined criteria including DC phenotype, viability, IL-12p70 production and the ability to stimulate MLR as metrics of whole oxidized tumor lysate-pulsed DC immunogenicity and functionality. Development and optimization of this unique method is now being tested in a clinical trial of autologous oxidized tumor lysate-pulsed DC in clinical-scale in recurrent ovarian, primary peritoneal or fallopian tube cancer (NCT01132014).

Figures

Figure 1

Comparison of the phenotype and IL-12p70 production of mature HOCl-oxidized lysate-loaded DCs generated in human AB serum-free or serum-containing AIM-V and CellGenix DC media, and supplementation with different concentrations of GM-CSF and IL-4. (A-B) DCs cultured in CellGenix DC media containing 2% human AB serum displayed a desirable phenotype with the highest level of DC-LAMP (*P = 0.005; one-way ANOVA) when compared to DCs cultured in other media conditions, and significantly higher CD86 (*P = 0.05) and lower CD14 (*P = 0.04) when compared to DCs cultured in serum-free CellGenix DC media. (C) DCs generated in CellGenix DC media plus 2% human AB serum produced slightly, though not significantly, higher amount of IL-12p70 compared to DCs generated in other media conditions (P = 0.2; one-way ANOVA). (D) 500 IU/ml recombinant GM-CSF and 250 IU/ml recombinant IL-4 were sufficient concentrations for generating a favorable immature DC phenotype with low CD14, and intermediate HLA-DR and CD1c. All the data presented were the mean of 6 independent experiments (i.e. DCs from 6 different individuals) in research-scale cultures in Nunclon™Δ Surface T25 cm2 flasks ± standard error of the mean (SEM). P values were determined with unpaired Student's t test unless otherwise stated. * denoted that P values were significant.

Figure 2

Comparison of the effect of Nunclon™Δ Surface, Corning® cell-culture surface and Corning® ultra-low attachment surface on DC phenotype and differentiation. DC were cultured with CellGenix DC media containing 2% human AB serum in T25 cm2 flasks that had the same surface properties as one of the three commercially available cell factories - Nunclon™Δ Surface 1-tray Cell Factories, Corning® CellSTACK® culture chambers or Corning® cell culture ultra-low attachment CellSTACK® chamber. DC phenotype was evaluated following lysate-loading and maturation for 16 h with LPS and IFN-γ. (A) DCs cultured in Nunclon™Δ Surface flasks (surrogate of Nunclon™Δ Surface 1-tray Cell Factories) exhibited an overall favorable immunophenotype with significantly higher DC-LAMP (**P = 0.003) and lower CD14 (*P = 0.01) when compared to DCs cultured in Corning® cell-culture surface flasks (surrogate of Corning® CellSTACK® culture chambers). Data were the mean of 6 independent experiments (i.e. DCs from 6 different individuals) in research-scale cultures ± SEM. (B) DCs that were cultured in Corning® ultra-low attachment surface (surrogate of Corning® ultra-low attachment cell factory chambers) expressed significantly lower CD86 (**P = 0.011) and HLA-DR (***P = 0.0004) following lysate-loading and maturation when compared to DCs cultured in Nunclon™Δ Surface (surrogate of Nunclon™Δ Surface 1-tray Cell Factories) and treated the same way. Data were the mean of 3 independent experiments (i.e. DCs from 3 different individuals) in research-scale cultures ± SEM. P values were determined with unpaired Student's t test. * denoted that P values were highly significant.

Figure 3

Comparison of TrypLE™ Select animal-free tyrpsin immobilization and cold PBS with cell scraping for harvesting mature HOCl-oxidized lysate-loaded DCs. (A) DCs were matured with LPS and IFN-γ for 16 h following lysate-loading. It was observed that mature lysate-loaded DCs that were harvested using TrypLE™ Select animal-free tyrpsin replacement exhibited approximately 30-40% reduction in the levels of MHC Class I (P = 0.13; NS) and HLA-DR (P = 0.15; NS) compared to DCs that were harvested with cold DPBS and cell scraping. Data were the mean of 3 independent experiments (i.e. DCs from 3 different individuals) ± SEM in research-scale Nunclon™Δ Surface T25 cm2 flasks. P values were determined with unpaired Student's t test. NS indicated that P values were not significant. (B) Mature lysate-loaded DCs that were harvested with TrypLE™ Select animal-free tyrpsin replacement were as efficient as DCs that were harvested with cold DPBS and cell scraping in presenting MHC Class I-restricted HER-2/neu and MART-1 peptides and stimulating HER-2/neu and MART-1 CD8+ T cells in intracellular IFN-γ assessments.

Figure 4

Comparing the phenotype of DCs previously cryopreserved in infusible cryomedia or human AB serum containing 10% DMSO. (A) Middle panel, DCs that had been cyropreserved with infusible cryomedia showed similar expressions of CD86 and CD1c as fresh DCs (first panel) and higher CD80, CD40 and MHC Class I levels after being thawed and replated for 24 h in fresh human AB serum-supplemented CellGenix DC media. Third panel, DCs that were cryopreserved with 90% human AB serum plus 10% DMSO exhibited similar level of CD80, CD40, MHC Class I and CD1c as fresh DCs. DCs cryopreserved in infusible cryomedia expressed approximately 50% lower levels of HLA-DR (**P = 0.001) compared to fresh DCs. On the other hand, DCs cryopreserved in human AB serum plus 10% DMSO media expressed approximately 80% lower levels of HLA-DR (**P = 0.004) compared to fresh DCs. Data were from a clinical-scale Nunclon™Δ Surface 1-tray Cell Factory and was representative of 3 independent research-scales experiments (i.e. DCs from 3 different individuals) in Nunclon™Δ Surface T25 cm2 flasks. (B) Summary of flow cytometry results from the mean of 3 independent research-scales experiments in Nunclon™Δ Surface T25 cm2 flasks (MFI ± SEM). P values are determined with unpaired Student's t test. ** denotes P value highly significant.

Figure 5

Time-course of IL-12p70 production by DCs. IL-12p70 was detected from fresh DCs from 6 h post LPS and IFN-γ stimulation and peaked at 16 h. Fresh replated DCs and thawed replated DCs continued to produce IL-12p70 albeit at a lower level than fresh DCs. Data were the mean of 3 independent experiments (i.e. DCs from 3 different individuals) in duplicates in research-scale Nunclon™Δ Surface T25 cm2 flasks ± SEM. IL-12p70 concentration is expressed as pg/ml.

Figure 6

Schematic diagram of optimized DC preparation.

Figure 7

Mixed leukocyte reaction capability of optimized DCs. DCs generated from clinical-scale Nunclon™Δ Surface 1-tray Cell Factory and DCs generated from research-scale Nunclon™Δ Surface T25 cm2 flasks using the optimized protocol depicted in Figure 6 were equally potent in stimulating robust mixed leukocytes reactions from 3 different individuals' T cells.