Despite improvements in therapy for acute myelogenous leukemia (AML), a significant percentage of patients still relapse and succumb to their disease. Dendritic cell immunotherapy offers the promise of potentially effective supportive therapy for a variety of neoplastic conditions; and the use of DCs loaded with tumor antigens is now recognized as an important investigational therapy. Though a variety of methods have been used to load DC vaccines, the loading of the MHC class II compartment with tumor lysate has predominated. The priming of a class II-mediated (CD4) T-cell response may be crucial to the success of DC immunotherapy as such a response is likely required for the development of memory CD8+ T-cells. DC cross-presentation is credited with the ability of lysate-loaded DCs to prime both CD4 and CD8 T-cell responses, enabling the generation of CD8+ CTLs without the loading of the MHC class I compartment (i.e. the cytoplasm). Recently, however, several reports have raised doubts as to the efficiency of cross-presentation as a mechanism for CTL priming in vivo. To examine this issue, we have loaded human DCs with both AML tumor lysate and mRNA. This technique allows the full repertoire of class I antigens to be presented without dependence upon cross-presentation; and, moreover, provides a full complement of class II antigens necessary for CD4 T-cell priming and the generation of memory responses.Methods: CD14+ precursors were isolated from normal donor PBPCs by magnetic separation. Immature DCs were then generated by culturing precursors for six days in GM-CSF and IL-4. Lysate was produced by three successive freeze/thaw cycles of blasts. mRNA was extracted from blasts using Trizol and oligo-dT separation. Immature DCs were pulsed for three hours with AML lysate and subsequently electroporated with AML mRNA. Loaded DCs were matured for 48 hours with IL-1β, TNF-α, IL-6, and PGE2 and then used to prime autologous T-cells. Short-term responses were assayed on day 5 of the 1st stimulation. Memory responses were assayed on day 10 of a tertiary stimulation.Results: Doubly-loaded DCs can prime a superior T-cell response in vitro in comparison to that of singly-loaded DCs, demonstrating a 30–70% increase in IFN-γ ELISpots over lysate-loaded DCs (p<0.001) and a 3–4 fold increase in ELISpots in comparison to mRNA loaded DCs (p<0.001). These results were verified by flow cytometry which showed 35% of CD8+ T-cells primed by doubly-loaded DCs were CD69+/IFN-γ+ vs. 14% of CD8+ T-cells primed by lysate-loaded DCs (p<0.001). This enhancement may be based upon both an upregulation of CD83 surface expression (p<0.0019) of doubly-loaded DCs and/or the upregulation of B7.1/B7.2 that accompanies elevated CD40L signaling. Memory responses were also greatly improved, with a 126% increase in total ELISpots (double loaded DCs versus lysate loaded DCs; p<0.03) and a 187% increase in total IFN-γ secretion (p<0.03). Unloaded (p<0.01) and mRNA (p<0.007) loaded DCs exhibited a virtual inability to generate memory T-cells in vitro, suggesting that the perpetuation of the memory response is reliant upon T-cell help.Conclusion: DCs doubly-loaded with lysate and mRNA are more efficient in the generation of primary and secondary immune responses than are singly-loaded DCs. The clinical administration of such doubly-loaded DCs may offer an important therapeutic option to patients with AML.
Read full abstract