Centrosome-mediated microtubule remodeling during axon formation in human iPSC-derived neurons.

Axon formation critically relies on local microtubule remodeling and marks the first step in establishing neuronal polarity. However, the function of the microtubule-organizing centrosomes during the onset of axon formation is still under debate. Here, we demonstrate that centrosomes play an essential role in controlling axon formation in human-induced pluripotent stem cell (iPSC)-derived neurons. Depleting centrioles, the core components of centrosomes, in unpolarized human neuronal stem cells results in various axon developmental defects at later stages, including immature action potential firing, mislocalization of axonal microtubule-associated Trim46 proteins, suppressed expression of growth cone proteins, and affected growth cone morphologies. Live-cell imaging of microtubules reveals that centriole loss impairs axonal microtubule reorganization toward the unique parallel plus-end out microtubule bundles during early development. We propose that centrosomes mediate microtubule remodeling during early axon development in human iPSC-derived neurons, thereby laying the foundation for further axon development and function.

Quantitative mapping of transcriptome and proteome dynamics during polarization of human iPSC-derived neurons.

The differentiation of neuronal stem cells into polarized neurons is a well-coordinated process which has mostly been studied in classical non-human model systems, but to what extent these findings are recapitulated in human neurons remains unclear. To study neuronal polarization in human neurons, we cultured hiPSC-derived neurons, characterized early developmental stages, measured electrophysiological responses, and systematically profiled transcriptomic and proteomic dynamics during these steps. The neuron transcriptome and proteome shows extensive remodeling, with differential expression profiles of ~1100 transcripts and ~2200 proteins during neuronal differentiation and polarization. We also identified a distinct axon developmental stage marked by the relocation of axon initial segment proteins and increased microtubule remodeling from the distal (stage 3a) to the proximal (stage 3b) axon. This developmental transition coincides with action potential maturation. Our comprehensive characterization and quantitative map of transcriptome and proteome dynamics provides a solid framework for studying polarization in human neurons.

Cortical anchoring of the microtubule cytoskeleton is essential for neuron polarity.

The development of a polarized neuron relies on the selective transport of proteins to axons and dendrites. Although it is well known that the microtubule cytoskeleton has a central role in establishing neuronal polarity, how its specific organization is established and maintained is poorly understood. Using the in vivo model system Caenorhabditis elegans, we found that the highly conserved UNC-119 protein provides a link between the membrane-associated Ankyrin (UNC-44) and the microtubule-associated CRMP (UNC-33). Together they form a periodic membrane-associated complex that anchors axonal and dendritic microtubule bundles to the cortex. This anchoring is critical to maintain microtubule organization by opposing kinesin-1 powered microtubule sliding. Disturbing this molecular complex alters neuronal polarity and causes strong developmental defects of the nervous system leading to severely paralyzed animals.

CAPS-1 requires its C2, PH, MHD1 and DCV domains for dense core vesicle exocytosis in mammalian CNS neurons.

CAPS (calcium-dependent activator protein for secretion) are multi-domain proteins involved in regulated exocytosis of synaptic vesicles (SVs) and dense core vesicles (DCVs). Here, we assessed the contribution of different CAPS-1 domains to its subcellular localization and DCV exocytosis by expressing CAPS-1 mutations in four functional domains in CAPS-1/-2 null mutant (CAPS DKO) mouse hippocampal neurons, which are severely impaired in DCV exocytosis. CAPS DKO neurons showed normal development and no defects in DCV biogenesis and their subcellular distribution. Truncation of the CAPS-1 C-terminus (CAPS Δ654-1355) impaired CAPS-1 synaptic enrichment. Mutations in the C2 (K428E or G476E) or pleckstrin homology (PH; R558D/K560E/K561E) domain did not. However, all mutants rescued DCV exocytosis in CAPS DKO neurons to only 20% of wild type CAPS-1 exocytosis capacity. To assess the relative importance of CAPS for both secretory pathways, we compared effect sizes of CAPS-1/-2 deficiency on SV and DCV exocytosis. Using the same (intense) stimulation, DCV exocytosis was impaired relatively strong (96% inhibition) compared to SV exocytosis (39%). Together, these data show that the CAPS-1 C-terminus regulates synaptic enrichment of CAPS-1. All CAPS-1 functional domains are required, and the C2 and PH domain together are not sufficient, for DCV exocytosis in mammalian CNS neurons.

SNAP-25 gene family members differentially support secretory vesicle fusion.

Neuronal dense-core vesicles (DCVs) transport and secrete neuropeptides necessary for development, plasticity and survival, but little is known about their fusion mechanism. We show that Snap-25-null mutant (SNAP-25 KO) neurons, previously shown to degenerate after 4 days in vitro (DIV), contain fewer DCVs and have reduced DCV fusion probability in surviving neurons at DIV14. At DIV3, before degeneration, SNAP-25 KO neurons show normal DCV fusion, but one day later fusion is significantly reduced. To test if other SNAP homologs support DCV fusion, we expressed SNAP-23, SNAP-29 or SNAP-47 in SNAP-25 KO neurons. SNAP-23 and SNAP-29 rescued viability and supported DCV fusion in SNAP-25 KO neurons, but SNAP-23 did so more efficiently. SNAP-23 also rescued synaptic vesicle (SV) fusion while SNAP-29 did not. SNAP-47 failed to rescue viability and did not support DCV or SV fusion. These data demonstrate a developmental switch, in hippocampal neurons between DIV3 and DIV4, where DCV fusion becomes SNAP-25 dependent. Furthermore, SNAP-25 homologs support DCV and SV fusion and neuronal viability to variable extents - SNAP-23 most effectively, SNAP-29 less so and SNAP-47 ineffectively.