EHD3 is localized on the tubular structures of early endosomes, and it regulates their trafficking pathway. However, the regulatory mechanism of EHD3-containing tubular structures remains poorly understood. An in vitro liposome co-sedimentation assay revealed that EHD3 interacted with phosphatidic acid through its helical domain and this interaction induced liposomal tubulations. Additionally, inhibiting phosphatidic acid synthesis with diacylglycerol kinase inhibitor or lysophosphatidic acid acyltransferase inhibitor significantly reduced the number of EHD3-containing tubules and impaired their trafficking from early endosomes. These results suggest that EHD3 and phosphatidic acid cooperatively regulate membrane deformation and trafficking from early endosomes.

Early endosomes (EEs) are known to be a sorting station for internalized molecules destined for degradation, recycling, or other intracellular organelles. Segregation is an essential step in such sorting, but the molecular mechanism of this process remains to be elucidated. Here, we show that actin is required for efficient recycling and endosomal maturation by producing a motile force. Perturbation of actin dynamics by drugs induced a few enlarged EEs containing several degradative vacuoles and also interfered with their transporting ability. Actin repolymerization induced by washout of the drug caused the vacuoles to dissociate and individually translocate toward the perinuclear region. We further elucidated that cortactin, an actin-nucleating factor, was required for transporting contents from within EEs. Actin filaments regulated by cortactin may provide a motile force for efficient sorting within early endosomes. These data suggest that actin filaments coordinate with microtubules to mediate segregation in EEs.

Phosphatidylinositol 4-kinase IIα (PtdIns4KIIα) localizes to the trans-Golgi network and endosomal compartments and has been implicated in the regulation of endosomal traffic, but the roles of both its enzymatic activity and the site of its action have not been elucidated. This study shows that PtdIns4KIIα is required for production of endosomal phosphatidylinositol 4-phosphate (PtdIns(4)P) on early endosomes and for the sorting of transferrin and epidermal growth factor receptor into recycling and degradative pathways. Depletion of PtdIns4KIIα with small interfering RNA significantly reduced the amount of vesicular PtdIns(4)P on early endosomes but not on Golgi membranes. Cells depleted of PtdIns4KIIα had an impaired ability to sort molecules destined for recycling from early endosomes. We further identify the Eps15 homology domain-containing protein 3 (EHD3) as a possible endosomal effector of PtdIns4KIIα. Tubular endosomes containing EHD3 were shortened and became more vesicular in PtdIns4KIIα-depleted cells. Endosomal PtdIns(4,5)P2 was also significantly reduced in PtdIns4KIIα-depleted cells. These results show that PtdIns4KIIα regulates receptor sorting at early endosomes through a PtdIns(4)P-dependent pathway and contributes substrate for the synthesis of endosomal PtdIns(4,5)P2.

A large GTPase dynamin, which is required for endocytic vesicle formation, regulates the actin cytoskeleton through its interaction with cortactin. Dynamin2 mutants impair the formation of actin comets, which are induced by Listeria monocytogenes or phosphatidylinositol-4-phosphate 5-kinase. However, the role of dynamin2 in the regulation of the actin comet is still unclear. Here we show that aberrant actin comets in dynamin2-depleted cells were rescued by disrupting of microtubule networks. Depletion of dynamin2, but not cortactin, significantly reduced the length and the speed of actin comets induced by Listeria. This implies that dynamin2 may regulate the actin comet in a cortactin-independent manner. As dynamin regulates microtubules, we investigated whether perturbation of microtubules would rescue actin comet formation in dynamin2-depleted cells. Treatment with taxol or colchicine created a microtubule-free space in the cytoplasm, and made no difference between control and dynamin2 siRNA cells. This suggests that the alteration of microtubules by dynamin2 depletion reduced the length and the speed of the actin comet.