Search

SUMMARYIntracellular organelles of mammalian cells communicate with each other during various cellular processes. The functions and molecular mechanisms of such interorganelle association remain largely unclear, however. We here identified voltage-dependent anion channel 2 (VDAC2), a mitochondrial outer membrane protein, as a binding partner of phosphoinositide 3-kinase (PI3K), a regulator of clathrin-independent endocytosis downstream of the small GTPase Ras. VDAC2 was found to tether endosomes positive for the Ras-PI3K complex to mitochondria in response to cell stimulation with epidermal growth factor and to promote clathrin-independent endocytosis as well as endosome maturation at membrane contact sites. With a newly developed optogenetics system to induce mitochondrion-endosome association, we found that, in addition to its structural role in such association, the pore function of VDAC2 is also required for the promotion of endosome maturation. Our findings thus uncover a previously unappreciated role of mitochondrion-endosome association in the regulation of endocytosis and endosome maturation.HighlightsThe mitochondrial protein VDAC2 binds PI3K and tethers endosomes to mitochondriaVDAC2 promotes clathrin-independent endocytosisVDAC2-PI3K interaction induces acidification of endosomes associated with mitochondriaThe pore function of VDAC2 also contributes to endosome maturation at contact sites

The oncogenic tyrosine kinase BCR-ABL activates a variety of signaling pathways and plays a causative role in the pathogenesis of chronic myelogenous leukemia (CML); however, the subcellular distribution of this chimeric protein remains controversial. Here, we report that BCR-ABL is localized to stress granules and that its granular localization contributes to BCR-ABL-dependent leukemogenesis. BCR-ABL-positive granules were not colocalized with any markers for membrane-bound organelles but were colocalized with HSP90a, a component of RNA granules. The number of such granules increased with thapsigargin treatment, confirming that the granules were stress granules. Given that treatment with the ABL kinase inhibitor imatinib and elimination of the N-terminal region of BCR-ABL abolished granule formation, kinase activity and the coiled-coil domain are required for granule formation. Whereas wild-type BCR-ABL rescued the growth defect in IL-3-depleted Ba/F3 cells, mutant BCR-ABL lacking the N-terminal region failed to do so. Moreover, forced tetramerization of the N-terminus-deleted mutant could not restore the growth defect, indicating that granule formation, but not tetramerization, through its N-terminus is critical for BCR-ABL-dependent oncogenicity. Our findings together provide new insights into the pathogenesis of CML by BCR-ABL and open a window for developing novel therapeutic strategies for this disease.Key words: BCR-ABL, subcellular localization, stress granule.

The discovery of fluorescent proteins (FPs) has revolutionized cell biology. The fusion of targeting sequences to FPs enables the investigation of cellular organelles and their dynamics; however, occasionally, such fluorescent fusion proteins (FFPs) exhibit behavior different from that of the native proteins. Here, we constructed a color pallet comprising different organelle markers and found that FFPs targeted to the mitochondria were mislocalized when fused to certain types of FPs. Such FPs included several variants of Aequorea victoria green FP (aqGFP) and a monomeric variant of the red FP. Because the FFPs that are mislocalized include FPs with faster maturing or folding mutations, the increase in the maturation rate is likely to prevent their expected localization. Indeed, when we reintroduced amino acid substitutions so that the FP sequences were equivalent to that of wild-type aqGFP, FFP localization to the mitochondria was significantly enhanced. Moreover, similar amino acid substitutions improved the localization of mitochondria-targeted pHluorin, which is a pH-sensitive variant of GFP, and its capability to monitor pH changes in the mitochondrial matrix. Our findings demonstrate the importance of selecting FPs that maximize FFP function.

B-cell translocation gene 1 and 2 (BTG1 and BTG2) are members of the BTG/Tob antiproliferative protein family, which is able to regulate the cell cycle and cell proliferation. We previously reported that BTG1, BTG2, Tob, and Tob2 are degraded via the ubiquitin-proteasome pathway. In this study, we investigated the mechanism of polyubiquitination of BTG1 and BTG2. Since the Skp1-Cdc53/Cullin 1-F-box protein (SCF) complex functions as one of the major ubiquitin ligases for cell cycle regulation, we first examined interactions between BTG proteins and components of the SCF complex, and found that BTG1 and BTG2 were capable of interacting with the SCF complex containing Cullin-1 (a scaffold protein) and Skp1 (a linker protein). As the SCF complex can ubiquitinate various target proteins by substituting different F-box proteins as subunits that recognize different target proteins, we next examined which F-box proteins could bind the two BTG proteins, and found that Skp2, β-transducin repeat-containing protein 1 (βTrCP1), and βTrCP2 were able to associate with both BTG1 and BTG2. Furthermore, we obtained evidence showing that βTrCP1 enhanced the polyubiquitination of both BTG1 and BTG2 more efficiently than Skp2 did, and that an F-box truncated mutant of βTrCP1 had a dominant negative effect on this polyubiquitination. Thus, we propose that BTG1 and BTG2 are subjected to polyubiquitination, more efficiently when it is mediated by SCFβTrCP than by SCFSkp2.

We investigated whether the assembly/disassembly of the 26S proteasome is regulated by phosphorylation/dephosphorylation. The regulatory complex disassembled from the 26S proteasome was capable of phosphorylating the p45/Sug1/Rpt6 subunit, suggesting that the protein kinase is activated upon dissociation of the 26S proteasome or that the phosphorylation site of p45 becomes susceptible to the protein kinase. In addition, the p45-phosphorylated regulatory complex was found to be incorporated into the 26S proteasome. When the 26S proteasome was treated with alkaline phosphatase, it was dissociated into the 20S proteasome and the regulatory complex. Furthermore, the p45 subunit and the C3/alpha2 subunit were cross-linked with DTBP, whereas these subunits were not cross-linked by dephosphorylating the 26S proteasome. These results indicate that the 26S proteasome is disassembled into the constituent subcomplexes by dephosphorylation and that it is assembled by phosphorylation of p45 by a protein kinase, which is tightly associated with the regulatory complex. It was also revealed that the p45 subunit is directly associated with the 20S proteasome alpha-subunit C3 in a phosphorylation-dependent manner.