The use of dendritic cells (DCs) in therapeutic cancer vaccination requires their loading with tumor-specific antigen(s). DEC-205, a phagocytosis receptor mediating antigen uptake, is associated with CD8+ T-cell responses in mice. Here we fused an anti–DEC-205scFv to an HLA-DP4–restricted epitope from the tumor antigen MAGE-A3, and examined the suitability and efficacy of DEC-205 to deliver a helper epitope to human monocyte-derived DCs (moDCs). The construct specifically bound DEC-205 on human moDCs without negative impact on DC phenotype and function. We measured antigen presentation with specific autologous CD4+ T cells, generated by TCR-RNA transfection. DEC-205 targeting resulted in significant major histocompatibility complex class II–restricted antigen presentation, and was superior to loading DCs by electroporation of mRNA encoding endosome-targeted MAGE-A3-DCLAMP or by direct peptide pulsing. Anti–DEC-205scFv-MAGE-A3 was presented 100 times more efficiently than the control constructs. DC maturation before or during incubation with anti–DEC-205scFv-MAGE-A3 reduced the interleukin-10/interleukin-2 ratio. Moreover, we successfully applied the DEC-205 targeting strategy to moDCs from malignant melanoma patients. Again, DEC-205–targeted mature DCs (mDCs) presented the antigen more efficiently than peptide-pulsed DCs and maintained their stimulatory capacity after cryoconservation. Thus, DEC-205 targeting represents a feasible and effective method to deliver helper epitopes to DCs in anticancer vaccine strategies, which may also be suitable for DC targeting in vivo.

Adoptive T-cell therapy relying on conventional T cells transduced with T-cell receptors (TCRs) or chimeric antigen receptors (CARs) has caused substantial tumor regression in several clinical trials. However, genetically engineered T cells have been associated with serious side-effects due to off-target toxicities and massive cytokine release. To obviate these concerns, we established a protocol adaptable to GMP to expand and transiently transfect γ/δ T cells with mRNA. PBMC from healthy donors were stimulated using zoledronic-acid or OKT3 to expand γ/δ T cells and bulk T cells, respectively. Additionally, CD8+ T cells and γ/δ T cells were MACS-isolated from PBMC and expanded with OKT3. Next, these four populations were electroporated with RNA encoding a gp100/HLA-A2-specific TCR or a CAR specific for MCSP. Thereafter, receptor expression, antigen-specific cytokine secretion, specific cytotoxicity, and killing of the endogenous γ/δ T cell-target Daudi were analyzed. Using zoledronic-acid in average 6 million of γ/δ T cells with a purity of 85% were generated from one million PBMC. MACS-isolation and OKT3-mediated expansion of γ/δ T cells yielded approximately ten times less cells. OKT3-expanded and CD8+ MACS-isolated conventional T cells behaved correspondingly similar. All employed T cells were efficiently transfected with the TCR or the CAR. Upon respective stimulation, γ/δ T cells produced IFNγ and TNF, but little IL-2 and the zoledronic-acid expanded T cells exceeded MACS-γ/δ T cells in antigen-specific cytokine secretion. While the cytokine production of γ/δ T cells was in general lower than that of conventional T cells, specific cytotoxicity against melanoma cell lines was similar. In contrast to OKT3-expanded and MACS-CD8+ T cells, mock-electroporated γ/δ T cells also lysed tumor cells reflecting the γ/δ T cell-intrinsic anti-tumor activity. After transfection, γ/δ T cells were still able to kill MHC-deficient Daudi cells. We present a protocol adaptable to GMP for the expansion of γ/δ T cells and their subsequent RNA-transfection with tumor-specific TCRs or CARs. Given the transient receptor expression, the reduced cytokine release, and the equivalent cytotoxicity, these γ/δ T cells may represent a safer complementation to genetically engineered conventional T cells in the immunotherapy of melanoma (Exper Dermatol 26: 157, 2017, J Investig Dermatol 136: A173, 2016).

Transfection with RNA is an attractive method of Ag delivery to dendritic cells (DCs), but has not yet been standardized. We describe in this study the methods to efficiently generate an optimized mature monocyte-derived DC vaccine at clinical scale based on the electroporation of several RNAs either into immature DC followed by maturation or, alternatively, directly into mature DCs, which has not been possible so far with such high efficiency. Electroporation of DCs resulted in high yield, high transfection efficiency (>90%), and high migration capacity. Intracellular staining allowed the study of the expression kinetics of Ags encoded by the transfected RNAs (MelanA, MAGE-3, and survivin) and a validation of the vaccine (≥90% transfection efficiency). Expression of all three Ags peaked 3–4 h after electroporation in DC transfected either before or after maturation, but decreased differently. The DC vaccine can also be cryopreserved and nevertheless retains its viability, stimulatory capacity as well as migratory activity. In addition, we uncover that DC transfected after rather than before maturation appear to be preferable vaccines not only from a production point of view but also because they appear to be immunologically superior for CTL induction in sharp contrast to common belief. DCs transfected after maturation not only more effectively generate and present the Mage-3.A1 and MelanA.A2.1 epitopes to T cell clones, but they even are superior in priming to the standard proteasome-dependent MelanA.A2.1 wild-type prototype tumor epitope, both in terms of T cell expansion and effector function on a per cell basis.

We explored the possibility to target Ewing's sarcoma family of tumors (ESFT) by redirecting T cells. To this aim, we considered NKG2D-ligands (NKG2D-Ls) as possible target antigens. Detailed analysis of the expression of MICA, MICB, ULBP-1, -2, and -3 in fourteen ESFT cell lines revealed consistent expression of at least one NKG2D-L. Thus, for redirecting T cells, we fused a CD3ζ/CD28-derived signaling domain to the ectodomain of NKG2D, however, opposite transmembrane orientation of this signaling domain and NKG2D required inverse orientation fusion of either of them. We hypothesized that the particularly located C-terminus of the NKG2D ectodomain should allow reengineering of the membrane anchoring from a native N-terminal to an artificial C-terminal linkage. Indeed, the resulting chimeric NKG2D receptor (chNKG2D) was functional and efficiently mediated ESFT cell death triggered by activated T cells. Notably, ESFT cells with even low NKG2D-L expression were killed by CD8^pos and also CD4^pos cells. Both, mRNA transfection and lentiviral transduction resulted in high level surface expression of chNKG2D. However, upon target-cell recognition receptor surface levels were maintained by tranfected RNA only during the first couple of hours after transfection. Later, target-cell contact resulted in strong and irreversible receptor down-modulation, whereas lentivirally mediated expression of chNKG2D remained constant under these conditions. Together, our study defines NKG2D-Ls as targets for a CAR-mediated T cell based immunotherapy of ESFT. A comparison of two different methods of gene transfer reveals strong differences in the susceptibility to ligand-induced receptor down-modulation with possible implications for the applicability of RNA transfection.