Myosin IIb controls actin dynamics underlying the dendritic spine maturation

Abstract

Precise control of the formation and development of dendritic spines is critical for synaptic plasticity. Consequently, abnormal spine development is linked to various neurological disorders. The actin cytoskeleton is a structural element generating specific changes in dendritic spine morphology. Although mechanisms underlying dendritic filopodia elongation and spine head growth are relatively well understood, it is still not known how spine heads are enlarged and stabilized during dendritic spine maturation. By using rat hippocampal neurons, we demonstrate that the size of the stable actin pool increases during the neuronal maturation process. Simultaneously, the treadmilling rate of the dynamic actin pool increases. We further show that myosin IIb controls dendritic spine actin cytoskeleton by regulating these two different pools of F-actin via distinct mechanisms. The findings indicate that myosin IIb stabilizes the stable F-actin pool through actin cross-linking. Simultaneously, activation of myosin IIb contractility increases the treadmilling rate of the dynamic pool of actin. Collectively, these data show that myosin IIb has a major role in the regulation of actin filament stability in dendritic spines, and elucidate the complex mechanism through which myosin IIb functions in this process. These new insights into the mechanisms underlying dendritic spine maturation further the model of dendritic spine morphogenesis.

Introduction

Dendritic spines are small protrusions from neuronal dendrites. During development, spine shape changes from long, thin, headless protrusions (dendritic filopodia) to spines with short, narrow necks and large bulbous heads (mushroom spines) (Hotulainen and Hoogenraad, 2010). Most of the post-synaptic terminals of excitatory synapses reside in the dendritic spines, and the spine morphology and size modifies the synapse function. Abnormal spine shape has been detected in many neurological diseases (Calabrese et al., 2006, van Spronsen and Hoogenraad, 2010). In order to detect and understand pathogenic changes leading to neurological diseases it is fundamental to know how spines develop normally. The actin cytoskeleton is a structural element that regulates changes in spine shape as well as shape maintenance. The beauty of the actin cytoskeleton as a building element is its capacity to treadmill actin monomers through the filaments, making the actin cytoskeleton tremendously dynamic. Actin binding proteins regulate the treadmilling rate in several ways; they can facilitate the incorporation of monomers onto the barbed ends of actin filaments, protect the filaments from depolymerizing factors or increase the rate of depolymerization of actin monomers from the pointed end of the filament (Le Clainche and Carlier, 2008). Based on the treadmilling rate, actin filaments in dendritic spines can be divided into a dynamic pool (time constant < 1 min) and a stable pool (time constant ~ 17 min) (in addition, to an ‘enlargement pool’) (Honkura et al., 2008, Star et al., 2002). Synapse activity affects the proportions of the dynamic and stable F-actin pools, as well as the actin treadmilling rate (Honkura et al., 2008, Okamoto et al., 2004, Star et al., 2002). An abnormal treadmilling rate causes abnormal shape and dynamics of the dendritic spines and is a plausible cause of Baraitser–Winter syndrome (Hotulainen et al., 2009, Okamoto et al., 2007, Rivière et al., 2012).

Recently, we and others have revealed mechanisms underlying dendritic filopodia elongation as well as spine head growth from headless protrusions (Hotulainen and Hoogenraad, 2010). However, it is not known how spine heads grow further and get stabilized during dendritic spine maturation. The current view is that spine maturation does not change actin dynamics (Star et al., 2002). Taking into account that dendritic spine morphology and dynamics change during maturation (Dunaevsky et al., 1999, Oray et al., 2006, Takahashi et al., 2003), and that the actin cytoskeleton is the major structural element underlying these changes (Hotulainen and Hoogenraad, 2010), we found it surprising that there would be no apparent change in actin dynamics or the proportions of dynamic and stable F-actin pools.

Interestingly, we found out that there is a significant increase in the relative size of the stable F-actin pool during neuronal maturation. Furthermore, we were interested to know how F-actin stability can be regulated in dendritic spines. It is known that cofilin-1 depletion or over-expression of CaMKII can increase the relative size of the stable F-actin pool in dendritic spines (Hotulainen et al., 2009, Okamoto et al., 2007). To find other players involved in the regulation of actin filament stability, especially during spine maturation, we turned to myosin IIb which has been shown to be important for the normal development and function of dendritic spines (Hodges et al., 2011, Rex et al., 2010, Ryu et al., 2006, Zhang et al., 2005) and is re-located into dendritic spines during neuronal maturation (Ryu et al., 2006). In contrast to its recently proposed function as a de novo actin nucleator in neurons (Rex et al., 2010), we show that myosin IIb over-expression increases the relative size of the F-actin stable pool in dendritic spine heads. In line with this, myosin IIb silencing by siRNA depleted the stable F-actin pool indicating a major role in the regulation of actin filament stability. The myosin IIb motor function was not required for actin filament stabilization; in contrast, active and contractile myosin IIb increased the treadmilling rate of the dynamic F-actin pool.