The balance of actin filament polymerization and depolymerization maintains a steady state network treadmill in neuronal growth cones essential for motility and guidance. Here we have investigated the connection between depolymerization and treadmilling dynamics. We show that polymerization-competent barbed ends are concentrated at the leading edge and depolymerization is distributed throughout the peripheral domain. We found a high-to-low G-actin gradient between peripheral and central domains. Inhibiting turnover with jasplakinolide collapsed this gradient and lowered leading edge barbed end density. Ultrastructural analysis showed dramatic reduction of leading edge actin filament density and filament accumulation in central regions. Live cell imaging revealed that the leading edge retracted even as retrograde actin flow rate decreased exponentially. Inhibition of myosin II activity before jasplakinolide treatment lowered baseline retrograde flow rates and prevented leading edge retraction. Myosin II activity preferentially affected filopodial bundle disassembly distinct from the global effects of jasplakinolide on network turnover. We propose that growth cone retraction following turnover inhibition resulted from the persistence of myosin II contractility even as leading edge assembly rates decreased. The buildup of actin filaments in central regions combined with monomer depletion and reduced polymerization from barbed ends suggests a mechanism for the observed exponential decay in actin retrograde flow. Our results show that growth cone motility is critically dependent on continuous disassembly of the peripheral actin network.

A growth cone is a motile structure at the tips of axons that is driven by the actin network and guides axon extension. Low actin adhesion to the substrate creates a stationary actin treadmill that allows leading-edge protrusion when adhesion increases in response to guidance cues. We use experimental measurements in the Aplysia bag growth cone to develop and constrain a simple mechanical model of the actin treadmill. We show that actin retrograde flow is primarily generated by myosin contractile forces, but when myosin is inhibited, leading-edge membrane tension increases and drives the flow. By comparing predictions of the model with previous experimental measurements, we demonstrate that lamellipodial and filopodial filament breaking contribute equally to the resistance to the flow. The fully constrained model clarifies the role of actin turnover in the mechanical balance driving the actin treadmill and reproduces the recent experimental observation that inhibition of actin depolymerization causes retrograde flow to slow exponentially with time. We estimate forces in the actin treadmill, and we demonstrate that measured G-actin distributions are consistent with the existence of a forward-directed fluid flow that transports G-actin to the leading edge.

5-HT promotes neurite growth via IP3-dependent Ca2+ release in neuronal growth cones. Outgrowth depends on increased rates of actin array treadmilling mediated by Ca2+-calcineurin–dependent cofilin activation. This mode of growth contrasts with substrate-dependent responses, for which retrograde actin flow and advance rates have been inversely correlated.

We showed recently that a complex between Duo1p and Dam1p is required for both spindle integrity and kinetochore function in the budding yeast Saccharomyces cerevisiae. To extend our understanding of the functions and interactions of the Duo1p/Dam1p complex, we analyzed the novel gene product Dad1p (for Duo1 and Dam1 interacting). Dad1p physically associates with Duo1p by two-hybrid analysis, coimmunoprecipitates with Duo1p and Dam1p out of yeast protein extracts, and shows interdependent localization with Duo1p and Dam1p to the mitotic spindle. These results indicate that Dad1p functions as a component of the Duo1p/Dam1p complex. Like Duo1p and Dam1p, Dad1p also localizes to kinetochore regions in chromosomes spreads. Here, we also demonstrate by chromatin immunoprecipitation that Duo1p, Dam1p, and Dad1p associate specifically with centromeric DNA in a manner that is dependent upon Ndc10 and partially dependent upon the presence of microtubules. To explore the functions of Dad1p in vivo, we generated a temperature-sensitive allele, dad1-1. This allele shows spindle defects and a mitotic arrest phenotype that is dependent upon the spindle assembly checkpoint. In addition, dad1-1 mutants undergo chromosome mis-segregation at the restrictive temperature, resulting in a dramatic decrease in viability.

PKC activation enhances myosin II contractility in the central growth cone domain while decreasing actin density and increasing actin network flow rates in the peripheral domain. This dual mode of action has mechanistic implications for interpreting reported effects of PKC on growth cone guidance and neuronal regeneration.