Characterization of NDE1 in T cells

Helper and cytotoxic T cells accomplish focused secretion through the clustering of vesicles around the microtubule organizing center (MTOC) and translocation of the MTOC to the target contact site, known as the immunological synapse (IS). Dynein has been implicated in both the processes. Here, using Jurkat cells as a model T cell system, we showed that the dynein-binding proteins NDE1 and p150 [superscript Glued] form mutually exclusive complexes with dynein and exhibit non-overlapping distributions in target-stimulated Jurkat cells. Immunofluorescent staining showed that NDE1, dynein and Lis1 accumulated at the immunological synapse. NDE1 was cloned from Jurkat cDNA and overexpressed in Jurkat cells as a fusion protein with N-terminal EGFP (EGFP-NDE1) tag. These cells were defective in MTOC polarization when activated by SEE-coated Raji B cells. However, Jurkat cells overexpressing NDE1 fused to a C-terminal monomeric EGFP (NDE1-mEGFP) tag did not show any defect in MTOC polarization. Immunofluorescent staining indicated that dynein failed to accumulate at the IS formed by activated Jurkat cells expressing EGFP-NDE1, but not by NDE1-mEGFP expressing cells. To confirm that NDE1 was required for MTOC translocation, we also used siRNA to deplete NDE1 in Jurkat cells. When NDE1 was depleted, dynein failed to accumulate at the IS and MTOC translocation was also greatly inhibited. Similar results were obtained when NDE1 was depleted in primary mouse CTLs. In this case, TcR transgenic OT-1 mouse CTL were transduced with lentiviruses expressing NDE1 shRNA. NDE1 depleted CTLs showed a reduction in MTOC translocation compared to untreated cells. While both of the C-terminal NDE1-GFP fusions accumulated at the IS, only NDE1-mEGFP showed the ability to bind dynein. There was no detectable coimmunoprecipitation of NDE1 and dynein when anti-GFP antibody was used to immunoprecipitate NDE1 or when anti-DIC antibody was used to immunoprecipitate the dynein intermediate chain. These observations lead to the conclusion that NDE1 accumulates at the IS independent of its binding to dynein. The ability of NDE1 to accumulated at the IS independent of dynein suggests that it is likely what anchors dynein at the IS. It also raises the question of how dynein accumulates at the IS. We observed reduction in NDE1 recruitment to the IS when Jurkat cells were treated with colchicine to depolymerize microtubules so recruitment of NDE1 likely requires microtubule motor for transport. Since dynein does not seem to be required, perhaps kinesin transports NDE1 to the IS. Finally, we investigated vesicle movements in Jurkat cells using expressed CTLA4-mCherry to selectively label secretory lysosomes, vesicles that cluster around the MTOC and accumulate at the central contact site of stimulated Jurkat-Raji cell pairs. Depletion of p150 [superscript Glued] prevented this vesicle accumulation but did not affect MTOC translocation. Immunoprecipitation data showing that NDE1 and p150 [superscript Glued] are in different dynein complexes, that they show different spatial distributions in the cell, and show no functional interdependence, we conclude that the NDE1/Lis1 and dynactin complexes separately mediate the two key components of T cell effector functions.