Axon branching in the developing vertebrate CNS
The function of the nervous system relies on the establishment of complex neuronal circuitry. During development, each neuron must connect with multiple synaptic targets, sometimes in widely divergent regions of the nervous system. For example, single cortical axons from the thalamus can ramify in the motor, somatosensory, and higher order sensory cortices . The generation of these intricate connections is crucial to the formation of highly interconnected networks of neurons. With organisms of larger size and increased complexity, this “problem of connectivity” becomes all the more challenging . One of the goals of neurosciences is to understand the strategies used to build these neuronal networks. In the case of neuronal connectivity, natural selection has adopted the strategy of axon branching: each neuron makes contacts with its multiple targets through branching of its axon (see Figure I-11). Therefore, understanding the mechanisms behind the control of axon branching is a critical step in the study of neuronal circuit development. Furthermore, in the adult nervous system, the formation of axon branches also plays a key role in the system response to injury and neurodegeneration, and might contribute to plasticity .
The understanding of the formation of axonal lamellipodia is still limited. Actin monomers in axonal waves undergo anterograde transport and turnover. Actin waves seem to propagate using directional actin treadmilling as the driving mechanical force . Actin polymerization/depolymerization is correlated with the presence of several actin binding proteins, such as Arp3, desphosphorylated cofilin (driving actin filament turnover during protrusive activity) or shootin within the actin waves , , . In cultured hippocampal neurons, doublecortin (DCX) accumulates both at the growth cones and within actin waves, and depletion of DCX prevents the formation of waves and axon branching. DCX can interact with microtubules and also with actin filaments. However neurons branching, which occurs through growth cone splitting, was not impaired by DCX depletion. To finish with, inhibition of myosin II, a motor driving actomyosin contractility along antiparallel filaments, has been shown to promote axon branching and the presence of actin waves .
Interaction between actin and microtubules
The fundamental mechanisms underlying axon collateral branching rely on interactions between dynamic microtubules and actin filaments. Indeed, timelapse imaging studies have shown that at branch points, the splaying of microtubules is colocalized with focal accumulation of actin filaments. Furthermore, application of drugs that repress either microtubules or actin dynamics also inhibit polymerization of the other cytoskeleton element , . Although understanding of the mechanism coordinating actin and microtubules reorganization is still rather limited, there are a few proteins that have attracted attention as candidates for the role of coordinator of both cytoskeletal systems during axon branching. First Drebrin , a F-actin-binding protein involved in actin filament bundling regulation, has been referred to as a link between the actin and microtubule system . Debrin binds to EB3.
Dynamic architectures of neuronal networks in vitro a microtubule plus end binding protein that associates with the tips of polymerizing microtubules . Debrin can be found in actin patches and in the proximal part of filopodia, an ideal position to guide the entry of microtubule plus ends into the filopodia shaft. Septins are a family of GTP binding proteins that regulate actin and microtubule organization, their crosstalk, and their binding to other effectors . Several studies have pointed out the combined role of Septin 6 and 7 for the coordination of the cytoskeleton during axon collateral branching –. Septin 6 is present in actin patches and recruits cortactin to trigger the formation of filopodia, while Septin 7 guides the entry of axonal microtubules into filopodia (see Figure I-18 b)).
NGF-induced branching and PI3K signaling pathway
The neurotrophin family consists of molecules that regulate neuronal growth and survival. Among them, NGF (Nerve Growth Factor) is one of the extracellular signals that mediates axon collateral branching , . Several studies have demonstrated that NGF promotes the formation of axon collateral branching in sensory axons through activation of the phosphatidylinositol-3 kinase (PI3K) pathway , . The PI3K pathway is a major regulator of axonal growth, through the control of gene expression and cytoskeleton elements during neuronal morphogenesis. PI3K is a lipid kinase that transforms PIP2 (phosphatidylinositol [4,5] biphosphate) into PIP3 (phosphatidylinositol [3,4,5] triphosphate). PIP3 can in turn recruit a wide range of proteins to the membrane. On the opposite, phosphatase such as PTEN can turn PIP3 back intro PIP2, and genetic deletion of axon branching in vivo increases branching . The current model for the NGF-signaling pathway has NGF activate the TrkA membrane receptor, which results in activation of the PI3K signaling pathway, thereby activating the RAC1 GTPase to drive WAVE1 activity which in turns activates the actin nucleating complex Arp 2/3 , as can be seen in Figure I-I-20.
Branching through growth cone splitting
The second mode of branching happens through growth cone splitting. Indeed, growth cone bifurcation at the tip of the extending axon can give rise to two axon branches. Axon branching through growth cone bifurcation is not considered to be a major mechanism for axon branching, and has been frequently linked to axon guidance . It is an essential mechanism to the development of the nervous system during embryogenesis. Nevertheless, in vertebrates, dorsal root sensory neurons have been reported to branch through growth cone bifurcation after entering into the spinal cord , , . The sensory axon growth cone splits into branches that project rostrally and caudally to their target fields. Growth cone bifurcation has also been studied in the context of axon guidance in C. elegans. . In vitro studies have described in details the different steps of branching through growth cone bifurcation. First, a suppression of protrusion at the leading edge of the axon is observed, in the direction corresponding to the axis of axon outgrowth. Meanwhile, the sides of the growth cone maintain their protrusive activity (see Figure I-21). This asymmetry results in the separation of the growth cone into distinct zones that eventually extend into mature branches.
Axon turning through growth cone steering
During the development of the nervous system, each neuron extends an axon that must navigate a complex environment to reach its target. The growth cone, which can be found at the tip of axons and dendrites, plays a crucial role in axon guidance. This highly dynamic structure can integrate multiple extracellular cues and respond to them by in a way that allows the axon to find its target with accuracy. In particular, according to the environmental directions it receives, the growth cone will continue its progress forward or decide to initiate turning in another direction, for example towards an attractive cue. The path of the growth cone in vivo is composed of adhesive molecules such as molecules from the extracellular matrix (for example fibronectin or laminin) or molecules presented on a neighboring cell (for example transmembrane cell adhesion molecules CAMs). On the other hand, anti-adhesive molecules surface-bound molecules (such as ephrins) can act as barriers to axon outgrowth by inhibiting growth cone advance. Finally, diffusible chemical cues (such as netrins) act as “roadsigns”, attractive or repulsive cues that guide the growth cone advance . In the following section, we will examine the different biological features that allow the growth cone to translate environmental guidance cues into steering. We will present the different steps of localized cytoskeletal remodeling that characterize growth cone turning.
Protrusion towards the selected site
As we saw before, dynamic remodeling of the actin cytoskeleton underlies growth cone motility. When a guidance cue is detected by the growth cone sensors, the actin cytoskeleton is asymmetrically reorganized to allow turning in the appropriate direction. In the case of an attractive cue, it is widely assumed that the response involves first increased protrusion on the side of the growth cone closest to the chemoattractant , as can be seen in Figure I-23. This could be due to a local attenuation of the actin retrograde flow resulting in an accumulation of actin on the attractive cue side. Marsick et al. have reported that application of chemoattractive guidance cues NGF or netrin 1 to dorsal root ganglion neurons induced increased protrusion, F-actin accumulation and increased barbed end density on the side closed to the gradient . On the opposite, in the case of a repulsive growth cone signal, turning results from retraction caused by local disruption of F-actin structures and actomyosin contraction on the side of the repulsive cue. These dynamic regulations of actin are mediated by multiple regulators, such as the actin nucleators, actin- related protein (Arp)2/3 complex and the formins, the F-actin polymerization factors (ENA/VASP), ADF/ cofilin which can promote actin depolymerization and motors such as myosin II which has been shown to drive actin retrograde flow along with actin treadmilling .
Site stabilization: engorgement and consolidation
After the protrusion step, microtubules proceed to invade the selected side in a process called engorgement. This polarized invasion of microtubules into the peripheral domain on one side of the growth cone is essential for it to turn towards the selected direction. This step requires bending and cutting of microtubules which is accomplished by several MAPs such as spastin and katanin . The engorgement step begins after actin finishes clearing from the corridor between the selected site and the central (C) domain, perhaps as F-actin behind the clutch is severed and removed (see Figure I-24). This is followed by the invasion of stable bundled C domain microtubules into the area of new growth, as a new region of the axon shaft is consolidated behind them, thereby fixing axon direction. This site-directed microtubule invasion might be regulated by the actin arcs situated in the T-domain . Finally, during the last stage of turning, microtubules are compacted by myosin II-containing actin arcs into the newly localized C domain. F-actin protrusive activity is suppressed in the new axon shaft region, which further promotes axon shaft consolidation .
Fabrication of photoresist stencil
The fabrication of a photoresist stencil was performed inside the IPGG clean room using the following protocol:
– Positive photosensitive resist (S1805 from Shipley) is spincoated on the silanized coverslips with an acceleration of 4000 rpm/s and a speed of 4000rpm during 30s,producing a resist layer of 0.5μm.
– The photoresist is then annealed over a heating plate at 115°C during 1min.
– The coverslip is exposed under UV light at a wavelength of 435nm (G-line) with a MJB4 aligner (Suss Microtech), through the chosen chrome mask, for an exposure time determined by the formula below.
– The exposed area of the photoresist are then dissolved through immersion in a developer solution (Megaposit MF-26A, Shipley) for 1min, during which the coverslip is gently agitated with a pair of tweezers.
– The sample is then rinsed in deionized water and dried with an air gun.
Table of contents :
Chapter I. Introduction
I.A. The nervous system
I.A.1. Structure and evolution of the nervous system
I.A.2. Cells in the nervous systems
I.A.3. The neuronal cytoskeleton
I.B. Axon branching in the developing vertebrate CNS
I.B.1. Collateral branching
I.B.2. Branching through growth cone splitting
I.C. Axon turning through growth cone steering
I.C.1. The cytoskeletal dynamics of turning
I.C.2. Signaling mechanisms underlying turning
I.D. Studying neuronal branching in vitro: a review of different systems
I.D.1. Studying neuronal cell development in vitro
I.D.2. Microengineering tools to study neuronal growth in vitro
I.D.3. Application of microengineering tools to the study of neuronal branching in vitro
Chapter II. Materials and Methods
II.B. Primary cell cultures
II.B.1. Neuronal cell cultures
II.C. Patterned substrates
II.C.1. Static patterns
II.C.2. Dynamic patterns
II.D. Microscopy observations
II.D.1. Time-lapse experiments
II.D.2. Fixed cells
II.E. Analysis methods
II.E.1. Growth cone analysis
II.E.2. Branching probability and measure of branch length
II.E.3. Time-lapse analysis
II.E.4. Statistical tests
Chapter III. Results
III.A. Static micropatterning to study neuronal branching
III.A.3. Analysis of growth cone morphology
III.A.4. Static analysis of branching
III.A.5. Dynamic analysis of neuronal branching on static patterns
III.B. Dynamic patterning to study neuronal branching
III.B.1. Context and objective of the project
III.B.2. Methodology and results
Chapter IV. Discussion
IV.A. Actin based exploration of the GC microenvironment
IV.B. Selection of direction by microtubules
IV.C. Axons prefer going straight