The strength of the mAb 20.1-induced stimulation obtained with BTN3A1-transduced CHO cells is remarkable and considerably higher than that observed with CHO Chr6 BTN3A1 cells (Fig. 1) or RAJI cells [8]. To some extent this might be due to the three- to fourfold higher BTN3A1 expression by CHO BTN3A1 compared with CHO Chr6 BTN3A1 cells,

but it cannot be excluded that Chr6 negatively affects this type of activation. Other cell type-specific features of the presenting cells may also act on reporter cell activation, e.g. BTN3A1-transduced murine L929 cells which, like CHO BTN3A1 cells, do not present PAg, induce a much weaker, although statistically significant, mAb 20.1-triggered response (data not shown). The strong mAb 20.1-induced stimulation observed with CHO BTN3A1 cells as presenters and Vγ9Vδ2 TCR53/4-CD28+ ubiquitin-Proteasome degradation T cells as reporter cells contrasts with Vavassori et al. [12], who found an excellent PAg- but not mAb 20.1-induced activation of Vγ9Vδ2 find more TCR-transgenic mouse reporter cells. Again differences between PAg- and mAb 20.1-induced activation could be due to differences between the presenting cells (human

A375 cells versus rodent cells), but could also be due to different reporter cell origin (cell line versus artificially in vivo-matured Vγ9Vδ2 T cells [12, 14]) and, finally, different TCR specificities might also play a role. Surprisingly, CHO Chr6 BTN3A1 cells, as well as several of the Chr6-containing hybridomas (Fig. 1), lacked the capacity to activate the reporter cells in the presence of the sec-butylamine. Alkylamines and aminobisphosphonates inhibit FPPS and therefore have been proposed to act by causing accumulation of IPP in the presenting cells [5]. Thus, either inhibition of FPPS activity is not the common feature that makes both substances Vγ9Vδ2 T-cell activators or alkylamines require additional cellular compound(s), which are missing in CHO Chr6 BTN3A1 cells, to exert FPPS inhibition. Independent evidence that, in addition to BTN3A1, human Chr6 is needed to convert rodent

cells into mediators of PAg-dependent Vγ9Vδ2 T cell activation was obtained using total human PBMCs as reporter cells. In this experiment we did not include mAb 20.1, since binding to BTN3A1 expressed by the γδ T cells would have complicated interpretation. these Figure 2 summarizes data comparing zoledronate-pulsed CHO Chr6 and CHO Chr6 BTN3A1 cells for induction of CD69 expression by the Vγ9Vδ2 T-cell population contained in total PBMCs. Most importantly, only the Vδ2+ T cells (essentially identical to Vγ9Vδ2 T-cell population) were activated. Furthermore, activation was better with CHO Chr6 BTN3A1 than with CHO Chr6 cells, while CHO (not shown) and CHO BTN3A1 cells failed to induce a PAg response. This experiment provides independent experimental proof that, in addition to BTN3A1, other gene(s) on Chr6 are indeed mandatory for PAg action.