Cultured Human B Cell APCs Expanded Using Il-4 and CD40 Ligand Preferentially Promote a Type 2 T Cell Response to Alloimmune Stimulation.

Abstract In vitro stimulation of human peripheral blood B cells with recombinant IL-4 and CD40 ligand (CD40L) markedly increases their expression of MHC and costimulatory molecules, thus enhancing antigenic peptide presentation to T cells. Because these cells proliferate extensively in vitro (unlike monocytes or dendritic cells), they represent a promising and convenient reagent for the generation and maintenance of antigen-specific T cells for use in a variety of experimental or therapeutic settings. However, the impact of this type of B cell APC on cytokine production by responder T cells has hitherto not been examined. To address this issue, we stimulated normal human T cells with either allogeneic B cells (generated in vitro) or with MNCs obtained from the same donor. After 7 days, T cells were washed and re-challenged with the same APCs. The resulting alloreactive cytokine response was measured using quantitative ELISPOT methods and expressed as the frequencies of IFN-γ, IL-4, and IL-5 producing cells per thousand responder cells added. B cell- and MNC-primed cell lines both produced vigorous lymphokine responses, but B cell-stimulated T cells consistently produced more IL-5 spots (mean of 265 vs. 98/1000 responders, p<0.002) and fewer IFN-γ spots (163 vs 386/1000 cells, p<0.005) than MNC-stimulated cells. Further, the ratio of IFN-γ to IL-5 spots was almost ten-fold lower in B cell-stimulated cultures compared to MNC-induced cultures (0.67 vs. 5.2, p<0.001). ELISPOT studies assessing the ratio of IFN-γ to IL-4 spots and ELISA assays comparing IFN-γ and IL-5 levels from culture supernatants demonstrated the same pattern of marked type 2 skewing by B cells. This pattern was unaffected by the presence of anti-IL-4 antibody suggesting type 2 skewing was not mediated by IL-4. Cytokine skewing produced by B cells or MNC could be partially reversed by swapping MNC and B cells during re-stimulation on day 7, but this plasticity was markedly reduced after 3 (weekly) cycles of B cell or MNC re-stimulation in vitro. Type 2 skewing by B cells was enhanced when monocytes were removed from responder T cell populations by either depleting CD14+ positive cells or by positive selection of T cells prior to stimulation. In contrast, type 2 polarization could be prevented using recombinant IL-12. Not all cells of B-cell origin share the same propensity to type 2 skewing observed with IL-4/CD40L-stimulated B cells; under identical conditions, EBV-transformed B cells stimulated alloimmune T cells to produce a strong type 1 cytokine response comparable to that produced by MNCs. In summary, IL-4/CD40L-stimulated B cells strongly promote a type 2 T cell response during primary alloimmune challenge; this skewing can become fixed after repeated B cell stimulation. Investigators using these cells as APC should be aware of this potential phenomenon, particularly during primary T cell responses. It is also important to consider the factors described above that may exacerbate or ameliorate this effect.