Mutations in genes essential for protein homeostasis have been identified in frontotemporal dementia (FTD) and amyotrophic lateral sclerosis (ALS) patients. Why mature neurons should be particularly sensitive to such perturbations is unclear. We identified mutations in Rab8 in a genetic screen for enhancement of an FTD phenotype associated with ESCRT-III dysfunction. Examination of Rab8 mutants or motor neurons expressing a mutant ESCRT-III subunit, CHMP2B Intron5 , at the Drosophila melanogaster neuromuscular junction synapse revealed synaptic overgrowth and endosomal dysfunction. Expression of Rab8 rescued overgrowth phenotypes generated by CHMP2B Intron5 . In Rab8 mutant synapses, c-Jun N-terminal kinase (JNK)/activator protein-1 and TGF-β signaling were overactivated and acted synergistically to potentiate synaptic growth. We identify novel roles for endosomal JNK-scaffold POSH (Plenty-of-SH3s) and a JNK kinase kinase, TAK1, in regulating growth activation in Rab8 mutants. Our data uncover Rab8, POSH, and TAK1 as regulators of synaptic growth responses and point to recycling endosome as a key compartment for synaptic growth regulation during neurodegenerative processes.

We have used clonal analysis and time-lapse video recording to study the proliferative behavior of purified oligodendrocyte precursor cells isolated from the perinatal rat optic nerve growing in serum-free cultures. First, we show that the cell cycle time of precursor cells decreases with increasing concentrations of PDGF, the main mitogen for these cells, suggesting that PDGF levels may regulate the cell cycle time during development. Second, we show that precursor cells isolated from embryonic day 18 (E18) nerves differ from precursor cells isolated from postnatal day 7 (P7) or P14 nerves in a number of ways: they have a simpler morphology, and they divide faster and longer before they stop dividing and differentiate into postmitotic oligodendrocytes. Third, we show that purified E18 precursor cells proliferating in culture progressively change their properties to resemble postnatal cells, suggesting that progressive maturation is an intrinsic property of the precursors. Finally, we show that precursor cells, especially mature ones, sometimes divide unequally, such that one daughter cell is larger than the other; in each of these cases the larger daughter cell divides well before the smaller one, suggesting that the precursor cells, just like single-celled eucaryotes, have to reach a threshold size before they can divide. These and other findings raise the possibility that such stochastic unequal divisions, rather than the stochastic events occurring in G 1 proposed by “transition probability” models, may explain the random variability of cell cycle times seen within clonal cell lines in culture.