Although phosphatidylinositol 5-phosphate (PtdIns5levels increase when cells are stimulated to move and we find PtdIns5to promote cell migration in tissue culture and in a model. expressing an oncogenic tyrosine kinase, high levels of PtdIns5were observed, linking this phosphoinositide potentially to cancer [14]. During infection with [15, 16] or [17], the injection of bacterial phosphoinositide TAK-375 phosphatases (IpgD and SopB), which produce PtdIns5from PtdIns(4,5)from PtdIns3via PtdIns(3,5)but also the TAK-375 structurally related PtdIns5[23]. Intriguingly, two other major hits in the migration screen were enzymes proposed to be involved in PtdIns5biogenesis, MTMR3, which is a phosphoinositide 3-phosphatase specific for PtdIns3and PtdIns(3,5)is thought to be formed by dephosphorylation of PtdIns(3,5)in cell migration. As the siRNA screen was carried out with pools consisting of four siRNAs, we verified results in the cell migration assay using the deconvoluted siRNAs for MTMR3, PIKfyve and FGD1 and compared them with control siRNA-treated cells (Fig 1D,E). All four different MTMR3 and FGD1 siRNAs caused a decrease in number of migrated cells, concordant with the knockdown on mRNA and protein level (supplementary Fig S1CCE online). For PIKfyve, individual siRNAs gave a relatively poor knockdown as reported previously [26]. To deplete PIKfyve efficiently, we therefore performed double knockdown using two siRNA oligonucleotides from the siRNA pool and could in this way observe a significant decrease in cell migration (Fig 1D,E). To verify the screening results further, we studied the role of PIKfyve and MTMR3 by live-cell imaging of FGF1-stimulated BJ fibroblasts in a wound healing assay. Knockdown of PIKfyve (Fig 1F) resulted in a significant decrease in cell migration velocity and TAK-375 persistence (Fig 1F). In addition, inhibiting PIKfyve with its specific inhibitor, YM201636, reduced cell velocity by 50% (Fig 1G). On depletion of MTMR3 (Fig 1F), BJ cells were unable to migrate into the wound and showed a significant decrease in velocity of 60% and a decrease in persistence (Fig 1F; supplementary Movie 1 online). Manipulating PIKfyve or MTMR3 in BJ cells did not cause any overt toxicity, which could potentially interfere with cell migration (supplementary Fig S2 online). Retransfection with siRNA-resistant EGFP-MTMR3 complementary DNA restored the control phenotype (Fig 1H; supplementary Movie 2 online). Interestingly, a point mutant of MTMR3 (C413S) without catalytic activity [24] was unable to rescue the knockdown phenotype (Fig 1H), indicating that MTMR3 acts enzymatically on phosphoinositides during cell migration. Depletion of PIKfyve or MTMR3 also inhibited the random migration of unstimulated cells (supplementary Fig S3 online). Together, these data indicate that both PIKfyve and MTMR3 are involved in cell migration. To explore the role of MTMR3 in directed cell migration, we monitored the ability of BJ cells to polarize and orientate towards a wound. The position of the Golgi can be used as a marker of cell polarization, as it reorients rapidly between the nucleus and the leading edge during directed cell migration [27]. On FGF1 stimulation, 86% of control cells oriented themselves CANPml towards the wound, whereas only 66% of MTMR3-depleted cells oriented successfully (Fig 2A,B). Consistent with this observation, we also observed that MTMR3-depleted cells were less perpendicular to the wound (Fig 2A). Quantification showed that 80% of control cells presented actin fibres perpendicular to the wound, whereas only 52% of the cells showed this on MTMR3 depletion (Fig 2C). This fits well with the observation that MTMR3-depleted cells showed decreased persistence, as cells unable to orientate their Golgi and cytoskeleton towards a signal are presumably incapable of undergoing persistent cell TAK-375 migration. Figure 2 Implication of PtdIns5in remodelling of the actin cytoskeleton. (A) For polarization studies, control or MTMR3 siRNA-treated cells were stained as indicated. The wound is located at the top of the images. (B) Quantification of cells oriented towards … In an attempt to further investigate the role of MTMR3 and PIKfyve in cell migration, we tested whether their depletion influences the trafficking of integrins. We were unable to detect any differences in 51 integrin recycling (supplementary Fig S4A online). Total and cell-surface levels of 5, 1 and 5 integrins were also unchanged (supplementary Fig S4B online). In addition, 5 integrin localization remained unchanged (supplementary Fig S4C online). The integrity of focal adhesions as measured by immunostaining with antibodies against vinculin, talin and FAK, was also unchanged on depletion of MTMR3 (supplementary Fig S4DCF online). Furthermore, TAK-375 to investigate the dynamics of focal adhesions, we tested the ability of depleted cells to spread on a fibronectin-coated surface, a process mediated by focal complexes. As shown in supplementary Fig S5G online, the knockdown cells spread similar to control cells. Interestingly, however, we observed a clear difference in the organization of.