A Common Progenitor for Macrophages, Osteoclasts and Dendritic Cells Newly Discovered in Human Bone Marrow and Cord Blood Yields Dendritic Cells with T-Cell Priming Ability

Dendritic cells (DC) have potent antigen-presentation and T-cell priming ability and therefore hold great promise in cancer immunotherapy. However, DC vaccination has not yet delivered a reliable clinical response rate, despite great efforts. Since primary DCs are rare (up 0.2-1.5% of circulating leukocytes), therapeutic DCs are generally derived from peripheral blood monocytes by culture with GM-CSF (moDC). Vaccines composed of moDC loaded with tumor antigens can induce potent and long-lasting tumour-specific immune responses in patients, but such positive results are infrequent and unpredictable. To improve success rate, research has focused on moDC culture regimens, antigen loading and activation strategies and methods of DC injection. Nevertheless, to date clinical trials using moDC have not yielded statistically significant treatment benefits over conventional strategies. Current attention has therefore shifted to the rare primary DCs that circulate in the blood under homeostatic conditions.

Knowing the identity of the precursors of these DCs may facilitate the ex-vivo or in-vivo generation of DCs via the homeostatic pathway, potentially yielding DCs with optimal T cell priming ability. We (Xiao et al. Stem Cell Rep. 2015) and others (Lee et al. J. Exp. Med. 2015) have recently identified a population with DC progenitor potential in human bone marrow and cord blood, respectively. This population can be isolated on basis of a CD34⁺ c-KIT⁺ FLT3⁺ IL3Rα^high phenotype and is furthermore Lin^- CD10^- CD11b^- CD45RA⁺ CD38⁺. We have shown that this population is highly enriched for or identical to a common progenitor (P) of macrophages (M), osteoclasts (O) and DCs (D) and termed it MODP. We also identified the progenitor directly upstream from the MODP that still has granulocyte (G) differentiation potential and termed it GMODP.

We hypothesized that DCs generated from GMODP or MODP under homeostatic conditions would have superb T-cell priming capacity. To examine this, the progenitors were sorted by flow cytometry from human bone marrow or cord blood and cultured with Flt3 ligand, M-CSF and IL-3 to generate DCs. We also tested the effect of a mensenchymal stem cell (MSC) feeder layer. Within 2-3 weeks of culture, 1000 DC progenitors generated approximately 150,000-250,000 DCs. Co-culture with MSC increased DC output significantly, at least 2 fold. The progenitor-derived DCs could be discerned into CD141⁺ conventional (c)DC, CD1c⁺ cDC and CD303⁺ plasmacytoid (p)DC. To study T-cell priming capacity of progenitor-derived DCs, we set up an in vitro DC-T co-culture assay. CD141⁺ cDC, CD1c⁺ cDC and CD303⁺ pDC were generated from GMODP or MODP of HLA-A2⁺ donors, flow cytometrically purified, activated with lipopolysaccharide and loaded with MART-1_26-35 peptide that represents a melanoma-derived tumor antigen. Primary T cells from peripheral blood of unrelated donors were retrovirally transduced to express a T cell antigen receptor (TCR) ab specific for the HLA-A2/MART-1_26-35 peptide complex. The ability of the DCs to prime a T-cell response was read out by antigen-specific CD8⁺ T cell proliferation. All DC subsets were able to induce MART-1 specific T cell proliferation, with the CD1c⁺ cDCs being most potent and the CD303⁺ pDC being least potent.

In conclusion: We have established a culture method to derive DCs with T-cell priming ability from a newly identified DC progenitor. These results are of value for improvement of DC-based immunotherapy.