BackgroundAntileukemic responses of immune reactive cells in AML-patients need to be improved. Combinations of blast-modulatory kitM (GM-CSF+PGE1) (vs control) convert myeloid blasts into dendritic cells of leukemic origin (DCleu), that effectively activate immune-cells against leukemic blasts. Exosomes are small (30–150 nm) membranous vesicles of endocytic origin produced by all cells under physiological and pathological conditions. Their involvement in nearly all aspects of malignant transformation has generated much interest in their biology, mechanisms responsible for information transfer and their role in immune-surveillance as well as -escape.Exosomes secreted by dendritic cells (DCs) have been shown to allow efficient activation of T lymphocytes, displaying potential as promoters of adaptive immune responses.Materials and Methods1)DC/DCleu-culture of blast containing AML patients’ whole blood (WB) (n=10) and of healthy volunteers(n=8) with kits, T-cell enriched mixed lymphocyte culture (MLC) with kit- vs un-treated WB, functional blast-cytotoxicity and, leukemia-specificity assays (Degranulation/intracellular cytokine-assays), Flowcytometric evaluation of blast-,DC- and lymphocyte composition before or after cultures. 2)Exosomes were isolated by immunoaffinity from serum, DC- and MLC-culture supernatants of 3 AML patients and 3 healthy volunteers. Exosomes were negatively stained and characterized by transmission electron microscopy (TEM). Fluorescence nanoparticle tracking analysis (fNTA) was performed to determine exosomal size and -concentration. Obtained results were compared in AML and healthy volunteers.ResultsAddition of kitM to blast-containing WB significantly increased frequencies of mature DC/DCleu and their subtypes compared to untreated WB without induction of blasts’ proliferation. Immune monitoring showed a continuous increase ofactivated/proliferating cells of the adaptive and innate immune system after Tcell-enriched MLC using kitM pretreated vs -untreated WB, suggesting a production/activation of (potentially leukemia-specific) cells after kit-stimulation. Moreover kit-pretreated WB regularly and significantly improved provision, activation as well as antileukemic and leukemia-specifically directed immune reactive cells after MLC. TEM showed exosome-like structures with a typically cup-shaped appearance without any differences between healthy and AML samples. fNTA revealed average vesicle sizes of 177±23 nm (healthy) and 178±17 nm (AML). Higher levels of EVs were detectable in AML samples compared to healthy controls in serum and after DC-culture, but lower levels after MLC independent of culture conditions.Interestingly, the number of EVs increased during cultivation of DC of AML and healthy samples, but not in AML-derived MLC samples.ConclusionsWe will provide data in AML patients and healthy volunteers about a potential role of DCs- and MLC-derived exosomes as biomarkers in immune responses, malignant progression or as potential therapeutic targets for AML patients.Disclosure InformationL. li: None. V. Mussack: None. E. Pepeldjiyska: None. A. Hartz: None. A. Rank: None. C. Schmid: None. E. Özkaya: None. S. Ugur: None. M. Pfaffl: None. H. Schmetzer: None.

were analysed using the CEL-seq2 platform. Interactions between tumour and immune cells were predicted using an unbiased ligand-receptor interaction analysis. Proteins secreted by tumour cells were analysed by performing mass spectrometry on conditioned medium from patient derived organoid cultures. The conditioned medium was concentrated using a 3kDa Millipore filter and prepared for LC-MS. Mass spectrometry data were acquired in data-dependent acquisition mode. For determining suppression, healthy donor PBMCs were stimulated with anti-CD28/anti-CD3 Dynabeads in the presence of recombinant proteins or concentrated conditioned medium from neuroblastoma organoids. In vitro killing assays were performed with GFP and luciferase transduced organoids and healthy donor PBMCs. Results Analysis of scRNA-seq of 25 neuroblastoma tumours showed a negative correlation between MIF and MDK expression and the cytotoxicity of NK cells, CD8 and gd-T cells within the tumour. These findings could be confirmed with a previously published SEQC bulk-RNAseq dataset containing 498 patients. Next to that, a higher expression of MIF and MDK correlated with poor survival in the same dataset. In the secretome from cultured neuroblastoma organoids, we have used mass spectrometry to identify MIF and MDK amongst the top 100 most abundant proteins from a total of ~1200 proteins. In vitro, we showed that rMIF and rMDK have a suppressive effect on the activation of T cells and the amount of cytotoxic factors such as granzyme B and Perforin are produced by the T cells. This confirms our hypothesis that MIF and MDK negatively influence the cytotoxicity of T cells. Conclusions Using two different unbiased analyses -scRNA-seq analysis of tumours and mass spectrometry of neuroblastoma organoid secretome-, we have identified MIF and MDK as immunosuppressive factor in neuroblastoma. Currently, we are testing several MIF and MDK inhibitors to test if T cell mediated killing of neuroblastoma can be increased if the immunosuppressive MIF and MDK are blocked.