Time-Lapse Super-Resolution Imaging and Optical Manipulation of Growth Cones in Elongating Axons and Migrating Neurons

Abstract

The growth cone is a highly motile tip structure that guides axonal elongation and directionality in differentiating neurons. Migrating immature neurons also exhibit a growth cone-like structure (GCLS) at the tip of the leading process. However, it remains unknown whether the GCLS in migrating immature neurons shares the morphological and molecular features of axonal growth cones and can thus be considered equivalent to them. Here, we describe a detailed method for time-lapse imaging and optical manipulation of growth cones using a super-resolution laser-scanning microscope. To observe growth cones in elongating axons and migrating neurons, embryonic cortical neurons and neonatal ventricular-subventricular zone (V-SVZ)-derived neurons, respectively, were transfected with plasmids encoding fluorescent protein-conjugated cytoskeletal probes and three-dimensionally cultured in Matrigel, which mimics the in vivo background. At 2-5 days in vitro, the morphology and dynamics of these growth cones and their associated cytoskeletal molecules were assessed by time-lapse super-resolution imaging. The use of photoswitchable cytoskeletal inhibitors, which can be reversibly and precisely controlled by laser illumination at two different wavelengths, revealed the spatiotemporal regulatory machinery and functional significance of growth cones in neuronal migration. Furthermore, machine learning-based methods enabled us to automatically segment growth cone morphology from elongating axons and the leading process. This protocol provides a cutting-edge methodology for studying the growth cone in developmental and regenerative neuroscience, being adaptable for various cell biology and imaging applications. Key features • Three-dimensional primary culture of migrating and differentiating neurons in Matrigel. • Visualization of fine morphology and dynamics of growth cones using super-resolution imaging. • Optical manipulation of cytoskeletal molecules in growth cones using photoswitchable inhibitors. • Machine learning-based extraction of growth cone morphology.

Keywords: Elongating axons; F-actin; Growth cone; Microtubules; Migrating neurons; Optical manipulation; Photoswitchable inhibitor; Postnatal neurogenesis; Super-resolution imaging; Ventricular–subventricular zone.

Figure 1.. Preparation of stripes on a 35 mm glass-bottom dish.

(A) Representative images showing Matrix 2C before and after corner cutting alongside a 35 mm glass-bottom dish. The corners of Matrix 2C are trimmed with a blade to fit within the glass-bottom dish. (B) Close-up views of the outlet channel and inlet slot (arrows) of the corner-cut Matrix 2C. (C) Top and bottom views of the dish with the corner-cut Matrix 2C attached. The stripe edge is marked on the glass surface as indicated by the arrow in the bottom view. (D, E) Critical steps for stripe preparation. A solution is applied to the inlet slot (D) and aspirated through the outlet channel using a P200 pipette. During the initial aspiration (step C4), the penetration of the solution into the stripe lanes is visible (D and E, boxed area).

Figure 2.. Subcellular distribution of cytoskeletal molecules in the growth cone of fixed migrating neurons.

Representative super-resolution images of ventricular–subventricular zone (V-SVZ)-derived cultured neurons stained for F-actin (green), tyrosinated tubulin (a marker for dynamic microtubules, red), and acetylated tubulin (a marker for stable microtubules, light blue). The nucleus is stained with Hoechst 33342 (blue). The growth cone is magnified in the right panels. In the growth cone, while F-actin and tyrosinated tubulin+ fibrous signals are observed in the central and peripheral domains, acetylated tubulin is observed only in the central domain. Scale bar, 5 μm.

Figure 3.. Dynamics of leading process growth cone on the Sdc2-coated stripe in the presence of CSPG.

(A) Experimental scheme. CSPG-rich Matrigel corresponds to 60% Matrigel/Ncan/L15 medium. (B–D) Representative super-resolution time-lapse images of ventricular–subventricular zone (V-SVZ)-derived cultured migrating neurons expressing Venus-CAAX (green) and DsRed (red), which label membranes and cytosol, respectively. While “minute-interval” imaging reveals the overall dynamics of the leading process growth cone (B), “second-interval” imaging visualizes the dynamics of fine cellular structures such as lamellipodia and filopodia. (C, D) Yellow, light-blue, and magenta arrows indicate formed, buried, and retracted filopodium, respectively. The structure and dynamics of lamellipodial and filopodial structures should be additionally analyzed by using actin probes such as GFP-actin or EGFP-UtrCH. Numbers indicate seconds from the first imaging frame (B–D). Scale bars, 5 μm (B–D). CSPG, chondroitin sulfate proteoglycan.

Figure 4.. Experimental scheme for optical manipulation of cytoskeletal molecules using photoswitchable inhibitors.

Following inhibitor addition to the media, the configuration and illumination ROI is quickly set. Laser illumination for inhibitor activation/inactivation and image acquisition are performed sequentially or simultaneously, depending on microscope specifications.

Figure 5.. Machine learning–based segmentation of growth cone area in axons and leading processes of cultured neurons.

(A) Training of segmentation of growth cone area using the Trainable Weka Segmentation plugin (FIJI). The axonal image is classified into growth cone (light blue), axonal shaft (red), and background (green). (B) Selection of training features in the Trainable Weka Segmentation plugin. Users should choose appropriate features by trial and error. (C, D) Successful segmentation of growth cone area in axons and leading process. Axonal and leading process growth cone areas are shown in blue (C) and dotted lines (D). Scale bars, 2 μm.