An Organotypic Slice Assay for High-Resolution Time-Lapse Imaging of Neuronal Migration in the Postnatal Brain

Summary

This protocol describes an organotypic slice assay optimized for the postnatal brain and high-resolution time-lapse imaging of neuroblast migration in the rostral migratory stream.

Abstract

Neurogenesis in the postnatal brain depends on maintenance of three biological events: proliferation of progenitor cells, migration of neuroblasts, as well as differentiation and integration of new neurons into existing neural circuits. For postnatal neurogenesis in the olfactory bulbs, these events are segregated within three anatomically distinct domains: proliferation largely occurs in the subependymal zone (SEZ) of the lateral ventricles, migrating neuroblasts traverse through the rostral migratory stream (RMS), and new neurons differentiate and integrate within the olfactory bulbs (OB). The three domains serve as ideal platforms to study the cellular, molecular, and physiological mechanisms that regulate each of the biological events distinctly. This paper describes an organotypic slice assay optimized for postnatal brain tissue, in which the extracellular conditions closely mimic the in vivo environment for migrating neuroblasts. We show that our assay provides for uniform, oriented, and speedy movement of neuroblasts within the RMS. This assay will be highly suitable for the study of cell autonomous and non-autonomous regulation of neuronal migration by utilizing cross-transplantation approaches from mice on different genetic backgrounds.

Protocol

I. Procedures

The following techniques should be performed under sterile conditions, in a laminar flow hood, using sterilized tools.

Preparation of glass bottom dishes for organotypic slices

1. Dishes must be prepared in sterile environment and using sterilized tools.
2. A 150μL drop of slice medium (see recipes) is placed in the center of the glass-bottom part of the dish with care to avoid air bubbles in the medium.
3. With a disposable syringe equipped with a 23 Gauge needle, multiple spots of glue (Rubber Cement - Elmer's, cat# E904) are placed on the square fringes adjacent to the circular coverslip that occupies the center of the culture dish, while leaving one side unglued for fluid exchange (Figure 1). The listed type of glue is not toxic to the slices or the cells as applied in this protocol. Care must be given to avoid placing any glue on the glass cover slip as this will impinge upon and obstruct imaging of the slices later. A nucleopore membrane (diameter 25mm, pore size 8.0μm - Whatman, cat# 110614) is placed on top of the glass coverslip, using the glue spots to secure it in place. This should be done using a micro forceps, while ensuring that air bubbles are not trapped between the glass coverslip and the membrane.
4. Add 1 mL slice medium on top of membrane. Dishes are placed in an incubator for 30 minutes and then on ice, until ready to use.

Extraction of early postnatal brains

The best results are obtained when slices are prepared from young postnatal mice (P1-P10).

1. Pups are terminally anesthetized (overdosed) by isofluorane or other approved methods. The head can be sprayed with 70% ethanol to increase sterility, followed by rapid decapitation using sharp scissors.
2. The head is stabilized by clamping the mandible with micro forceps. The skin is excised longitudinally from the neck to the snout. The skull is cut longitudinally and anteriorly starting at the cisterna magna, by making 1 medial and 2 lateral cuts (one on each side - Figure 2A). Care should be taken to minimize contact with underlying cortical tissue as the cranial flaps are removed away from the brain.
3. To improve stability of the tissue during vibratome sectioning, the lateral-most aspects of the brain are removed by making two sagittal cuts. The caudal aspect of the brain is also removed by making a cut at the rostral base of the cerebellum (Figure 2B).
4. The two hemispheres are separated by making a smooth cut along the midline fissure, and the two hemispheres are carefully scooped out of the skull and placed, medial surface down, in an embedding mold (Figs. 2C-D).

Sectioning of the host brain

1. The two hemispheres in the embedding mold are immediately covered with melted 3% low melting point DNA-grade agarose gel (Fisher, cat# BP1360-100) dissolved in tissue preparation buffer that is maintained at 37°C (see recipes). After 2 minutes of stabilization on a flat horizontal surface to ensure even hardening of the agarose, the molds are positioned on ice to complete the setting.
2. Once the gel containing the hemispheres is set, it is removed from the mold and trimmed, leaving 2-3mm of gel around the brain tissue.
3. The gel-embedded tissue is then mounted on the specimen disc of the vibratome, with the medial surface up, and secured with cyanoacrylate adhesive (Krazy glue or equivalent). Care must be taken to apply only minimal amounts of glue, as too much will have toxic effects on slices and migrating cells. Too much glue on the sides of the block will also impede the cutting, causing potential damage in the tissue.
4. The disc is installed in the vibratome specimen tray filled with ice-cold tissue preparation medium.
5. The tissue is sectioned at 150μm thickness, with the vibratome speed set on a slow-to-medium range (will depend on the vibratome used, and thus must be determined independently for optimal results). In our hands, the vibration frequency is optimal when set at maximum. The first few slices may be discarded.
6. As soon as the RMS-containing slices are released from the blade (the RMS is visible by the naked eye as a grey U-shaped structure extending from the SEZ to the OB), they are carefully teased out of the gel mold, and scooped using a small flathead spatula. The slices are placed on the nucleopore membrane of the glass-bottom dishes containing slice medium (Figure 3A-B). A typical early postnatal mouse brain will yield approximately 2-3 RMS sections per hemisphere when cut at 150 μm. It is critical that handling the slices is kept to a minimum as they are very fragile. The dishes with the slices are then transferred to an incubator.

Donor brain sectioning and RMS transplantation

1. Donor brains (brains expressing fluorescent reporters in migrating neuroblasts) are sectioned at 250μm thickness and slices are collected in ice-cold tissue preparation buffer.
2. Slices are immediately placed under a dissecting microscope with epifluorescence capability, and the RMS is gently microdissected, using micro forceps. One forceps is used to stabilize the slice while the other is used to make small cuts around the RMS until it is released from the slice. The excised RMS is then cut into small explants (approximately 200-500 μm in diameter) prior to transplantation.
3. Glass-bottom dishes containing host slices positioned on the nucleopore filters are removed from the incubator and placed under the dissecting microscope. Using visible light, the RMS is clearly identified and a small incision is made in the initial segment of the RMS.
4. Using a pipettor equipped with a 20 μL tip, a single donor RMS explant is transferred to the incised site in the host RMS. The explant is gently pushed into the incision to establish contact between the 2 tissues. To ensure this contact is stable, the explants are pushed slightly in between the slice and the nucleopore membrane.
5. Once all host slices with the RMS are transplanted, the dishes are returned to the incubator for at least 1 hour to allow the sections to settle on the membrane. Neuroblasts should begin to migrate from the explant into the host RMS after approximately 1-2 hours.

Time-lapse imaging of neuronal migration

1. The glass-bottom dishes are transferred from the incubator to the incubated chamber on the microscope (Figure 3C). Fluorescent neuroblasts can be imaged at intervals ranging from 0 to 10 minutes depending on the type of analysis desired. For example we image cytoskeletal dynamics during the migratory cycle of individual neuroblasts by setting our intervals at less than or equal to 2.5 minutes. However, population dynamics such as orientation and speed of migration are better captured at intervals ranging from 5 to 10 minutes. The choice of objectives is highly variable depending on the brand of microscope. For a Nikon C1 confocal microscope, we find that the 20x dry lens (Nikon Pan Fluor, N.A. 0.75, W.D. 0.35mm) is most suitable for our analyses. For best results on this confocal system, the pinhole is opened to a medium size (60μm diameter). For most appropriate migratory behavior, imaging is confined to cells deep within the thickness of the RMS, at least 20 μm away from either of the cut surfaces of the slice. Laser power should be optimized so that it is at a minimum, but that the details of individual neuroblasts remain visible.
2. Once the imaging is completed slices can be fixed with ice cold, and freshly prepared 4% paraformaldehyde, immunostained, and mounted on slides for further imaging. Since we do not coat our membranes with any adherent substrates such as laminin, the slices usually float away from the membrane once they are immersed in the staining buffers. Sections are imaged while maintaining their 150 μm thickness. It is also possible to cryopreserve and freeze the slices and use a cryostat to obtain thinner sections of the original slices. However, this will result in increase the incidence of artifacts in staining as well as alter the integrity of the tissue.

II. Materials/equipment

Preparation of glass bottom dishes for organotypic slices

• Small syringes (1 mL)
• 23-gauge needles
• Nucleopore Track-Etch Membrane - diameter 25mm, pore size 8.0μm - Whatman, cat# 110614
• Glass Bottom Culture Dishes - 35mm petri dish, 14mm Microwell, No. 1.5 coverglass - MatTech, cat# P35G-1.5-14-C
• Rubber Cement - Elmer's, cat# E904
• Basal Medium Eagle - Gibco, cat# 21010
• 1M Hepes (pH7.4)
• 1M D-glucose
• 100mM CaCl₂
• 100mM MgSO₄
• 1M NaHCO₃
• dH₂O
• 200mM L-glutamine
• Penicillin-streptomycin

Brain extraction and embedding

• Anesthetic (isofluorane, etc.)
• Microwave oven
• Low Melting Agarose - Fisher, cat# BP1360-100
• Krazy glue - cat#KG585
• Peel-A-Way Disposable Embedding Molds (R-40) - 22mmx40mm rectangular, 20mm deep - Polysciences, cat# 18646C

Brain sectioning and RMS transplantation

• Vibratome - Leica VT1000S and all accessory components for slice preparation
• 10X Hank's Balanced Salt Solution - Gibco, cat# 14185
• Microdissecting forceps #5 - Roboz, cat# RS-4976
• Microspatula - Fisher, cat# 21-401-15
• Stereomicroscope

Time-lapse imaging of organotypic slices

• Humidified Incubator, 5% CO₂
• Inverted Microscope equipped with incubator chamber and long distance working objectives (NA of 0.6 or higher)

III. Recipes

Buffer solution for tissue dissection and slice preparation (tissue preparation buffer)

Stock solution	Volume	Final Concetration
10X HBSS	50 mL	1X
1M Hepes (pH 7.4)	1.25 mL	2.5mM
1M D-Glucose	15 mL	30mM
1M CaCl2	0.5 mL	1mM
1M MgSO4	0.5 mL	1mM
1M NaHCO3	2 mL	4mM
dH2O	430.75 mL

Filter sterilize with a 0.2 μm filter and store at 4°C.

Culture medium for organotypic slices, tissue transplantation and imaging (slice medium)

Stock solution	Volume	Final Concentration
Basal Medium Eagle	35 mL
Tissue preparation buffer	12.9 mL
1M D-Glucose	1.35 mL	20mM
200mM L-glutamine	0.25 mL	1mM
Penicillin-streptomycin	0.5 mL	100units/ mL penicillin and 0.1mg/ mL streptomycin

Filter sterilize with a 0.2 μm filter and store at 4°C.

Preparation of Low-melting point agarose gel

Low-melting-point agarose is diluted in tissue preparation buffer at 0.3g/ mL in a 50 mL conical tube (see recipes). The tube is microwaved in increments of 5-10 seconds at high-power. Number of increments is dependent on total volume; for 10 mL, three increments (10-8-5 seconds, each) should suffice. The tube cap is carefully unscrewed between heating increments to release air pressure and avoid explosion of the tube. Caution must be taken as the tube content will be very hot. Once the agarose is completely dissolved, the tube is maintained in a 37°C water bath for at least 5 minutes to allow the temperature to stabilize prior to use. Prolonged exposure to room temperature will harden the gel. Although this must be avoided as much as possible, hardened gel can be reheated and re-melted for immediate use within 24 hours of initial preparation.

Immunohistochemistry on organotypic slices

Following imaging on the confocal microscope, slices may be fixed overnight at 4°C with 4% formaldehyde in PBS. Sections are then blocked overnight at 4°C, in 10% goat serum with 1% Triton X (Sigma, Cat. # S26-36-23) in PBS followed by overnight incubation with primary antibodies at 4°C. Fluorescently-tagged goat secondary antibodies are used for visualization (all diluted 1:1000, 1 hour incubation at room temperature). Labeled slices are thoroughly washed 5-6 times with ice cold PBS prior to mounting on glass slides and coverslipping.

IV. Representative Results

Our organotypic slice culture protocol has been thoroughly tested and optimized over that last few years for consistency in migration pattern and orientation. Analysis of cells emigrating from explants obtained from mice in which the expression of the red fluorescent protein, Td-tomato, is induced under the Nestin promoter (Nestin-Td tomato), reveals highly oriented and rapid migration of tdTomato+ neuroblasts into the host RMS (Figure 4A). High-magnification time-lapse analysis illustrates excellent resolution of the entire length of a migrating neuroblast during a 20-minute imaging session (Figure 4B).

Slices with donor Td-tomato+ cells were fixed and immunostained for different cellular components within the RMS. GFAP+ astrocytes and CD31+ blood vessels were revealed using fluorescent immunohistochemistry (Figure 5A). High-magnification analysis of slices stained for the cytoskeletal proteins actin and tubulin reveal non-uniform expression of these components by a cell in the midst of migration (Figure 5B).

Antibodies used in these examples: rabbit anti-RFP (Abcam, 1:250), rabbit anti-GFAP (Dako, 1:1000), rat anti-CD31(BD Pharmigen, 1:100), mouse anti-actin (Santa Cruz, 1:500), rabbit anti-tubulin (Sigma, 1:1000), goat anti-mouse Cy3 (Chemicon, 1:1000), goat anti-rabbit AlexaFluor 647 (Invitrogen, 1:1000), goat anti-rat AlexaFluor 488 (Invitrogen, 1:1000), goat anti-rabbit AlexaFluor 488 (Invitrogen, 1:1000).

Figure 1. Preparation of glass-bottom dishes for organotypic slices. Multiple spots of glue are placed around the circular glass-bottom component of the dish (red), leaving one side open for exchange of medium from underneath the filter. A 150μL drop of slice medium is placed in the center of the glass coverslip. A nucleopore membrane (blue) is then applied, shiny side down, while ensuring that no air bubbles are trapped between the glass coverslip and the membrane. One milliliter of slice medium (grey) is spread on top of membrane, and dishes are incubated at 37° C prior to use.

Figure 2. Brain extraction and preparation for sectioning. (A) The skull is exposed by incising the scalp from the neck to the snout (dotted line along the midline). The skull is then cut longitudinally and anteriorly starting at the cisterna magna, by making 1 medial and 2 lateral cuts (one on each side; 2A). (B) The lateral-most aspects of the cortex and the caudal aspect of the CNS are resected to improve stability of the tissue during vibratome sectioning. (C-D) The two hemispheres are then separated and placed medial face down in an embedding mold prior to application of 3% agarose gel dissolved in tissue preparation buffer.

Figure 3. Brain sectioning and cross-transplantation. (A) The host tissue is sectioned at 150μm thickness, and the RMS-containing sections are carefully positioned flat on the nucleopore membrane of cold glass-bottom dishes. (B) Donor brains (brains expressing fluorescent reporters in the RMS) are sectioned at 250μm thickness, and the slices are collected in ice-cold tissue preparation buffer. The donor RMS is microdissected and cut into small explants. Using a pipettor equipped with a 20 μL tip, individual RMS explants are transferred to an incised site in the host RMS. (C) After 1-2 hours of incubation, dishes are transferred to an incubated stage on a confocal microscope and migration is captured using time-lapse imaging. The photomicrograph is a representative low-magnification image of a typical slice (grey) set up 1 hour after transplantation (red explant from tdTomato+ RMS of a donor mouse; red dotted lines outline the RMS in the host slice).

Figure 4. Migration of neuroblasts from explants into the host RMS. (A) Nestin-tdTomato+ neuroblasts (red) migrate from the explants into the host RMS (dotted green line) 1 hour after transplantation. tdTomato+ cells invading the RMS of the host organotypic slices move in a highly oriented and rapid manner away from the SEZ and toward the OB. (B) The migratory cycle can be observed in high-power time-lapse images of a neuroblast over approximately a 20 minute period. Scale bar = 10 μm.

Figure 5. Immunohistochemical assessment of organotypic slices. Explanted neuroblasts (red) were fixed in the midst of migration 12 hours post transplantation. (A) Fluorescent immunohistochemical staining of the slice reveals a dense pool of GFAP+ astrocytes (blue) and scattered CD31+ blood vessels (green) within the RMS of the host slice. (B) The cytoskeleton of an isolated tdTomato+ migrating cell (red) in a host RMS is revealed by co-immunostaining using antibodies against actin (blue) and tubulin (green).

Discussion

Neuronal migration in the RMS is an essential component of postnatal neurogenesis in the olfactory bulbs ¹. Migration through the RMS occurs in a plane tangential to the surface of the brain. Tangentially migrating neuroblasts are distinct from radially migrating cells based on the location of their progenitor source, as well as the divergent fate of their final neuronal products ^{1, 2, 3}. The relatively pure population of tangentially migrating cells in the postnatal RMS makes this anatomically definable region an optimal platform for studying the mechanisms of tangential migration. Deciphering the cellular and molecular mechanisms of neuronal migration is critical for understanding many neurodevelopmental diseases (e.g., Ref. ⁴). This knowledge may further facilitate the identification of mechanisms that underlie the formation and spread of brain tumors (e.g., Ref. ⁵), and is essential for directing reprogrammed neuroblasts to sites of injury or neurodegeneration in future neuronal replacement strategies.

Despite enormous progress in the last few years, the current understanding of genetic and molecular mechanisms underlying neuronal migration, and their link to specific subcellular events remains fragmented. Much of the difficulty in making headways in this pursuit is due to the reliance of most migration studies on tissue fixation and immunohistochemistry, which have been used to draw indirect inferences regarding mechanisms of migration in the RMS and other migration streams in the embryonic brain. The 'fixed' approach is suboptimal since it only provides a snapshot of a cell, whereas temporal resolution is a key requirement for understanding a highly dynamic process such as cellular migration. Some insightful studies have addressed mechanisms of cellular migration using an in vitro method whereby explants, similar to those used in our protocol, are plated on plastic or glass substrates coated with components of the extracellular matrix (e.g.,Refs. ^{6, 7}). Although studies utilizing this assay have identified a number of mechanisms with clear roles in migration, the relevance of their findings to in vivo conditions is unclear. More recently, direct in vivo imaging of neuroblasts through the use of fibered confocal microscopy and MRI have been conducted ^8-10. However, such techniques yield low resolution images and thus are insufficient in subcellular assessment of neuronal migration. Another approach, the whole mount preparation devised by the Alvarez-Buylla group ¹¹, has been utilized to generate important physiological insights in the regulation of the subependymal stem cell niche ^12-14. Yet, this elegant approach cannot be utilized to study neuroblast migration in the RMS. Previous reports utilizing organotypic slices for analysis of migration in the embryonic brain ¹⁵ and the postnatal RMS have been published ^16-18. We have perfected this approach and provide here the protocol which allows for high-throughput recording of orientation, speed, and intra- and intercellular dynamics of migration in the RMS at high resolution.

The presented protocol poses some challenges that require practice, patience, and dedicated time to both the preparation and the analysis of results. The following suggestions will assist with carrying out this protocol:

• In our experience, mice in the age range of P1 to P10 provide the best results. Age-matching between donor and recipient brains also improves the reproducibility of the experiments.
• Since most reagents and culture dishes must be freshly prepared, our slice assay requires several hours of uninterrupted preparation and imaging in a single day. Once the brains are harvested consequent steps must be accomplished as fast as possible.
• Between decapitation and preparation of slices, all steps should be performed using ice cold reagents, and tissue manipulation should be kept to a minimum in order to avoid anatomical disruptions.
• The tools utilized for these dissections should be sterile and reserved for live tissue manipulations only. Never utilize tools that were ever in contact with fixatives such as formaldehyde.
• The slices are generally viable for up to 36 hours, with a visible decline in migration speed and orientation between 24 and 36 hours. Beware of this limitation when planning your experiments and analyses.

To our knowledge, the presented organotypic slice assay offers the best current approach for deciphering genetic and molecular mechanisms of neuroblast migration from the SEZ to the OB. Given the emerging information regarding the expression of signaling networks in the RMS ¹⁹, future experiments to link the identified signaling cascades to neuronal migration will massively benefit from our assay. Our protocol will facilitate the identification of candidate signaling molecules which may regulate neuroblast migration autonomously, while also allowing investigators to assess the role of chemokines and secreted factors in the regulation of migration in a non-autonomous manner, as we have recently demonstrated ^{18, 20}. Our presented technique can easily be modified for study of stem cell transplantation, dendritic/axonal outgrowth, and a whole host of other topics in development neurobiology and adult neurogenesis.