The Mouse Hindbrain As a Model for Studying Embryonic Neurogenesis

The mouse embryo forebrain is the most commonly employed system for studying mammalian neurogenesis during development. However, the highly folded forebrain neuroepithelium is not amenable to wholemount analysis to examine organ-wide neurogenesis patterns. Moreover, defining the mechanisms of forebrain neurogenesis is not necessarily predictive of neurogenesis in other parts of the brain; for example, due to the presence of forebrain-specific progenitor subtypes. The mouse hindbrain provides an alternative model for studying embryonic neurogenesis that is amenable to wholemount analysis, as well as tissue sections to observe the spatiotemporal distribution and behavior of neural progenitors. Moreover, it is easily dissected for other downstream applications, such as cell isolation or molecular biology analysis. As the mouse hindbrain can be readily analyzed in the vast number of cell lineage reporter and mutant mouse strains that have become available, it offers a powerful model for studying the cellular and molecular mechanisms of developmental neurogenesis in a mammalian organism. Here, we present a simple and quick method to use the mouse embryo hindbrain for analyzing mammalian neural progenitor cell (NPC) behavior in wholemount preparations and tissue sections.


Introduction
During embryonic mammalian development, NPCs divide in the ventricular (VZ) and subventricular (SVZ) zones of the expanding neuroepithelium to generate new neurons in a process termed 'neurogenesis'. The generation of new neurons and their NPC precursors has been investigated extensively in the forebrain 1 , whilst less is known about this process in other regions.
The forebrain is a complex and intricately folded structure that is largely studied with histological methods after tissue sectioning, which makes understanding neurogenesis patterns across the whole organ challenging. In addition, studies of forebrain neurogenesis are not necessarily predictive of neurogenic behavior in other brain regions or in the spinal cord. For example, signaling cues may elicit differing responses across distinct CNS regions, as observed in the case of ciliary neurotrophic factor and leukemia inhibitory factor, which promote self-renewal of NPCs in the lateral ganglionic eminence, but drive differentiation of spinal cord progenitors 2 . Furthermore, a sub-class of NPCs that populate the developing forebrain and contribute significantly to expansion of the cerebral cortex 1 are absent from the hindbrain and spinal cord 3 . Inversely, it is conceivable that the spinal cord and hindbrain contain alternative NPC subtypes not present in the cortex.
The mouse embryo hindbrain is the evolutionary oldest region of the mammalian brain and generates the cerebellum and brainstem. In spite of its conservation across species, relatively little is known about hindbrain neurogenesis, including NPC subtypes or their regulation. The majority of hindbrain research in the mouse has focused on the process of tissue segmentation, driven by Hox genes 4 , and the patterning of post-mitotic neurons 5 . In addition, the hindbrain has been used as a model for studying the mechanisms of developmental angiogenesis 6 .
In contrast to the mouse hindbrain, the zebrafish hindbrain has been used extensively to follow NPC differentiation and lineage progression in a vertebrate model organism (e.g., 7,8 ). The chick hindbrain has also been employed to study neurogenesis during vertebrate development (e.g., 9,10 ). Similar to the zebrafish hindbrain 11 , the chick hindbrain can be live imaged to study NPC behavior and regulation over time 12 .
Analogous longitudinal observation by live imaging is not presently possible in mammalian organisms, because they develop in utero. Moreover, targeted manipulation through techniques such as electroporation can be readily applied to free-living zebrafish embryos or chick embryos in ovo (e.g.,

Wholemount Immunofluorescence
1. If required, rehydrate hindbrains in serially decreasing dilutions of methanol in PBS (e.g., 75% methanol, 50% methanol, 25% methanol) at room temperature (RT) for 5 min each and then transfer to PBS. NOTE: A graded series of methanol is required to gently rehydrate hindbrains and ensure proper tissue preservation. 2. Permeabilize hindbrains for 30 min at 4 °C in PBS containing 0.1% Triton X-100 (PBT) with gentle agitation. 3. Incubate hindbrains for 1 h at 4 °C in PBT containing 10% heat-inactivated goat serum with gentle agitation. NOTE: Use serum from the host species that the secondary antibodies were raised in. For primary antibodies raised in goat, incubate in serum-free protein block, for example 5% bovine serum albumin in PBS or a suitable commercial alternative (see Table of Materials). This will reduce non-specific staining frequently observed when using primary antibodies raised in goat. 4. Incubate hindbrains overnight at 4 °C in PBT containing 1% heat-inactivated serum and primary antibodies with gentle agitation (e.g., rabbit anti-phospho histone H3 [pHH3] diluted 1:400). NOTE: For primary antibodies raised in goat, use PBT without serum. 5. Wash the hindbrains at 4 °C 5x with PBT for 1 h each. 6. Incubate hindbrains overnight at 4 °C in PBT containing appropriate fluorophore-conjugated secondary antibodies (e.g., goat anti-rabbit Alexa Fluor 488) at 1:200 in PBT targeted against the primary antibodies used. Keep hindbrains in the dark from here on to protect fluorophores from photobleaching. NOTE: For primary antibodies raised in goat, use anti-goat Fab fragments of secondary antibodies to reduce non-specific staining. 7. Wash the hindbrains at 4 °C 5x with PBT for 1 h each. 8. Postfix the hindbrains in 4% formaldehyde in PBS for 15 min at RT for long term preservation of antibody binding. Briefly rinse twice in PBS. 9. Cover a glass microscope slide with two layers of black electrical tape and excise a small square from the layered tape to create a pocket large enough to hold one hindbrain. 10. Transfer each hindbrain into a pocket with a Pasteur pipette, remove excess liquid and add an appropriate antifade reagent to the pocket before covering it slowly with a glass coverslip to avoid trapping air bubbles under the coverslip. Seal the coverslip and affix it to the slide with a thin layer of nail polish. Store slide at 4 °C in the dark until image acquisition (Protocol section 7).

Vibratome Sectioning and Floating Section Immunofluorescence
1. If required, rehydrate hindbrains from methanol as described in step 4.1.
2. Embed hindbrains in molten 3% w/v agarose prepared in distilled water. NOTE: Allow molten agarose to cool to approximately 55°C briefly before embedding to prevent heat damage to hindbrain.

Image Acquisition
1. Image samples using an epifluorescent or confocal laser-scanning microscope equipped with lenses suitable for aqueous media-mounted slides and optical filters suitable for the fluorophores used for immunolabeling. 2. To image the whole hindbrain or a hindbrain section, use a lens for 10x magnification (e.g., Figure 2B); to visualize hindbrains areas for quantification, use a 40X magnification (e.g., Figure 2D) and to visualize individual cells, use a lens with a 63x magnification (e.g., Figure  2C).

Alternative Methods
1. Following step 3.8, homogenize unfixed hindbrains to produce a single cell suspension for flow cytometry applications 15 . 2. Following step 3.8, homogenize unfixed hindbrains to extract RNA from single cell suspensions for RT-qPCR (e.g., 16 ). NOTE: Ensure all reagents and equipment are kept sterile and RNase-free throughout to prevent degradation of RNA. 3. Following step 3.8, homogenize unfixed hindbrains to isolate NPCs and propagate them in vitro as neurospheres for analysis of hindbrain NPC behavior 17 . NOTE: Ensure all reagents and equipment are kept sterile to prevent bacterial/fungal contamination of neurosphere cultures.

Representative Results
This section illustrates examples of results that can be obtained when studying neurogenesis in the mouse embryonic hindbrain through wholemount and tissue section analysis.
We show that wholemount immunolabeling of the microdissected hindbrain with an antibody for the mitotic marker pHH3 visualizes dividing NPCs in the VZ (Figure 2B -D). We show pHH3 + NPCs at a high magnification to highlight different stages of mitosis ( Figure 2C). We have illustrated that this labeling method is suitable to be performed across several consecutive stages of hindbrain development to observe the time course of NPC mitosis in this organ ( Figure 2D).
We show that imaging transverse immunolabeled vibratome sections of the hindbrain 1 h after EdU injection, visualizes the cleavage orientation of mitotic NPCs (Figure 3B), the pseudostratified, interkinetic nuclear migration pattern of cycling progenitors 18 (Figure 3B, D), and the overall VZ structure (Figure 3B -D). Note that mitotic pHH3 + NPCs are present only at the ventricular surface and not more basally (Figure 3C), which contrasts the basal division pattern of more committed NPCs in the forebrain 19 .
We also illustrate how cycling NPCs and their differentiated progeny can be labeled with BrdU or EdU to assess NPC lineage progression (Figure 4). The immunolabeling of transverse cryosections of the mouse hindbrain 1 day after BrdU injection for BrdU and Ki67 demonstrates the number and positioning of cycling NPCs in the neuroepithelium (Figure 4A, B), whereby the number of self-renewing NPCs can be defined by calculating the percentage of Ki67 + BrdU + cells amongst all BrdU + cells (Figure 4A, C).
Finally, we show an example of immunolabeling of hindbrain vibratome sections for RC2, an antigen in the neural-specific intermediate filament nestin, to visualize NPC endfeet ( Figure 5B) and processes ( Figure 5C). Vibratome sections, rather than thin cryosections, allow improved observation of the highly branched NPCs and also of overall neuroepithelial structure.

Discussion
This protocol describes how to use the mouse embryonic hindbrain as a model to study the mechanisms of developmental neurogenesis. Using a variety of different immunolabeling methods, hindbrain NPCs can be visualized and their number quantified in tissue sections or across organ wholemounts. The ease of dissection and flat anatomy ensures that the hindbrain can be imaged in an 'open book' preparation to gather information on organ-wide neurogenesis patterns.
We further show that NPC morphology and cell cycle-related NPC positioning can be easily visualized in floating-or cryosections of the hindbrain. Both behaviors may be exploited to define new progenitor populations, as has previously been performed in the mammalian telencephalon 20 . For example, early-formed Sox2 + neuroepithelia and Pax6 + apical radial glia are present in the hindbrain 21,22 , but the hindbrain lacks Tbr2 + basal progenitors 3 .
The protocol described here can also be adapted to observe the behavior of specific NPC subpopulations by fluorescent labeling for live imaging and/or lineage tracing in fixed tissues. This can be achieved, for example, by studying hindbrains from mice carrying the tamoxifen-inducible Sox1-iCreERT2 transgene and the Rosa26 tdTomato reporter 23 .
In addition to enhancing knowledge of murine neurogenesis, studying the hindbrain may elucidate broadly relevant neurogenic mechanisms that are shared across species, because the hindbrain is a highly conserved brain region that is expected to be more similar between vertebrate species than the forebrain.
As hindbrain neurogenesis takes place over a comparatively shorter time window than forebrain neurogenesis 23 , it is important to consider the need for comparing adequately staged embryos. Accordingly, experimental bias is avoided by counting and recording the number of somite pairs in an embryo prior to isolating its hindbrain. The hindbrain tissue itself is fragile and forceps should therefore be handled carefully when separating the hindbrain tissue from the head mesenchyme and meninges; a couple of 'practice runs' might therefore be advisable before dissection of valuable embryos is attempted. Furthermore, hindbrains should be transferred from one tube to another using a wide-bore Pasteur pipette rather than with forceps to avoid damage (the pipette's opening can be widened by cutting of the tip with clean scissors). Finally, although uncommon, the extent of EdU/BrdU incorporation into S-phase NPCs may be variable, in particular during short (i.e., 1 h) pulses. To improve labeling, ensure that EdU/BrdU is dissolved properly before loading the solution into the syringe and inject carefully into the peritoneal cavity. Poor injections, such as those made subcutaneously by accident, will result in losing or trapping the EdU/BrdU solution and prevent it from entering the circulation.

Disclosures
None of the authors have competing interests or conflicting interests.