Dissection, Culture and Analysis of Primary Cranial Neural Crest Cells from Mouse for the Study of Neural Crest Cell Delamination and Migration

Over the past several decades there has been an increased availability of genetically modified mouse models used to mimic human pathologies. However, the ability to study cell movements and differentiation in vivo is still very difficult. Neurocristopathies, or disorders of the neural crest lineage, are particularly challenging to study due to a lack of accessibility of key embryonic stages and the difficulties in separating out the neural crest mesenchyme from adjacent mesodermal mesenchyme. Here, we set out to establish a well-defined, routine protocol for the culture of primary cranial neural crest cells. In our approach we dissect out the mouse neural plate border during the initial neural crest induction stage. The neural plate border region is explanted and cultured. The neural crest cells form in an epithelial sheet surrounding the neural plate border, and by 24 h after explant, begin to delaminate, undergoing an epithelial-mesenchymal transition (EMT) to become fully motile neural crest cells. Due to our two-dimensional culturing approach, the distinct tissue populations (neural plate versus premigratory and migratory neural crest) can be readily distinguished. Using live imaging approaches, we can then identify changes in neural crest induction, EMT and migratory behaviors. The combination of this technique with genetic mutants will be a very powerful approach for understanding normal and pathological neural crest cell biology.


Introduction
The neural crest (NC) lineage is a transient, multipotent and migratory population of cells that appears exclusively in vertebrates during early embryonic development 1,2 . Neural crest derivatives are extremely diverse, and include glia, smooth muscle, melanocytes, neurons and craniofacial bone and cartilage 3,4 . Because the neural crest contributes to the function of many organ systems, this lineage is essential for human embryogenesis. Aberrant NC development is implicated in a wide range of the most common human birth defects (i.e., cleft lip and palate) 5 , and also disorders such as Hirschsprung's disease (HSCR), Wardensburg syndrome (WS), CHARGE syndrome and Williams Syndrome 6,7,8,9 . NC development has been explored in a number of non-mammalian model systems including Xenopus, chick and zebrafish models. In mammals, work in mouse models has identified some of the key genetic events underlying neural crest development; however, it has been more difficult to follow the cell biology of neural crest migration, due to the inaccessibility of the mouse embryo (reviewed elsewhere 10,11 ). Furthermore, while studies in chick, Xenopus and zebrafish have established a gene regulatory network for NC, loss of function studies in these animal models sometimes do not exhibit a comparable phenotype in mouse. For example, in Xenopus, zebrafish and chick, non-canonical Wnt signaling is one of the cellular mechanisms that allows the NC to acquire its migratory capacity 12,13,14,15 . However, in mouse, loss of non-canonical Wnt signaling does not seem to affect migration 16 . As in vivo NC migration has been difficult to track for long periods in mouse, it is unclear whether these species-differences reflect differing modes of migration, or differences in molecular regulation.
As noted, NC studies in mouse have been very challenging because the ex utero culture of embryos is laborious. Moreover, the NC is constantly in intimate contact with adjacent tissues such as mesoderm and neurectoderm. Recent use of neural crest-specific Cre drivers or exogenous dyes has allowed us to fluorescently label the migratory NC; however, these approaches are still limited. Despite multiple reports describing different techniques to visualize NC migration It is clear that there is a need for techniques that allow the handling and characterization of mammalian NC. We focused our efforts on the mouse cranial NC as it is the primary model for studying human craniofacial development and neurocristopathies. We refined our approach based on several interesting reports describing primary culture of NC cells 19,20,21 . Here, we thoroughly describe the optimal culture techniques for explanting primary NC cells. We demonstrate the live cell imaging method and the optimal use of different matrices to coat the culture plates. Our protocol describes how to capture the migration of live NC cells using an inverted microscope, which is intended as a guideline for use with other microscopes, as well as a detailed summary of our cellular analyses.
The expected result from the explant should be a beautifully laid out distribution of cells that are clearly distinguished under the microscope, where one can see three different populations of cells which represent (i) neural plate, (ii) premigratory, and, (iii) migratory neural crest cells. We demonstrate how to analyze the cell behaviors at the border of the premigratory population of cells during the epithelial-mesenchymal transition. We also focused our effort on studying fully migratory cells for cell speed, distance and cell morphology.

Protocol
All animal work has undergone ethical approval by the King's College London Ethical Review Process and was performed in accordance with UK Home Office Project License P8D5E2773 (KJL).
1. Dissection of the cranial neural plate is restricted to embryos at 8.5 days post coitum (dpc). Select embryos at the 5-8 somite stage. Dissect the uterus into PBS and cut the mesometrium to separate each embryo ( Figure 1A). The muscular wall of the uterus contracts and the decidual tissue will become visible ( Figure 1B). NOTE: Maintain embryos in the uterus in ice-cold PBS while dissections are performed one embryo at a time. Move embryos with a glass Pasteur pipette into fresh sterile PBS to improve visibility and reduce contamination. 2. Slide forceps between the muscle layer and decidual tissue and remove the muscle layer with a second pair of forceps ( Figure 1C). 3. Using forceps, pierce the deciduum at the edges of the mesometrial pole and with a second pair of forceps tear to open perpendicularly to the pole. 4. Peel back the decidual tissue with the forceps to visualize the Reichert's membrane. 5. Remove Reichert's membrane carefully. The visceral yolk sac becomes visible and the embryo can be seen inside ( Figure 1D). 6. Remove the visceral yolk sac and the amnion ( Figure 1E) and position the embryo in order to visualize the head fold ( Figure 1F). 7. Cut the head fold above the heart and scrape away the underlying mesoderm using forceps and/or eyelash tools to obtain a clean neural plate (NP) (Figure 1H). NOTE: The NP can be kept whole or divided down the anteroposterior axis so that each side can be plated individually. The neural plate border can be further trimmed away from the neural plate in order to minimize neuronal contributions to the explants. 8. Use a glass Pasteur pipette to transfer the dissected neural plate onto a hydrogel-coated dish filled with conditioned neural crest media. 9. Gently swirl the dish to position the NP in the middle of the well. This is important for maximizing the phase quality for live-cell imaging (on day 2). 10. Incubate overnight (or to desired time-point) at 37 °C in 5% CO 2 . Neural crest cells should be visibly migrating out of the neural plate.
NOTE: Cells usually attach within 6-8 h. After the explant attaches, allow more time to visualize migrating cells. Usually by 24 h post explant, we can find three distinguishable populations of cells. The first population, at the center of the explant, is the neural plate (NP). The second population, the premigratory NC (pNC), surrounds the NP in an epithelial sheet of cells. The third population, in the outside ring, is formed of migratory NC (mNC), which are bigger in size, and appear fully mesenchymal (Figure 2).

Day 2: Live cell imaging of murine cranial neural crest cells
NOTE: Imaging should be performed at 24 h post explanting to optimally image and quantify neural crest cell migration. NC induction media does not need to be refreshed before live cell imaging. Access to an inverted microscope, with a motorized stage and an incorporated environment chamber is required. Use multi-well tissue culture dishes suitable for imaging (Table of Materials).

Microscope set-up
1. Set the environment chamber at 37 °C and 5% CO 2 .
2. Pierce a hole into the lid of the tissue culture plate lid to allow the CO 2 needle, connected to the CO 2 humidification chamber, to sit within the plate. 3. Place the tissue culture dish into the specimen holder and tape down the plate lid and CO 2 needle to prevent shaking during multi-well acquisition. 4. Switch on the microscope controller, the stage controller and the imaging software. 5. Focus on the cranial NC cells at 10x magnification (with matching phase ring in the condenser selected). 6. Set high quality phase-contrast on the microscope by adjusting the field iris diaphragm, aperture iris diaphragm and centering telescope, as specified in the microscope set-up manual.
2. Phase contrast live-cell imaging 1. Set the directory or file location where the time-lapse files will be saved. 2. Set the exposure time, binning, and camera area. 3. Set the number of time points, duration of imaging and time interval between frames. 1. To quantify NC cell migratory capacity, set the microscope to 10x magnification, taking 1 frame every 5 min (217 time points over 18 h). To quantify cell morphology, set magnification to 40x, taking 1 frame/min (61 time points over 1 h). To quantify lamellipodial dynamics, set magnification to 40x or 60x magnification, taking 1 frame every 10 s (over 10 min).
4. For multi-well imaging, set the mechanical stage to move between selected XY positions of interest. Confirm that the cranial NC cells are in focus and the stage positions are correct. 5. Use the Acquire command to start time-lapse imaging. 6. Once time-lapse imaging is complete, review the multidimensional data and export .stk files for analysis. NOTE: .stk is a TIFF stack file. 7. Exit the software, shut down the computer and turn off the stage, camera and microscope controllers.

Imaging analysis: quantification of neural crest cell migration
NOTE: To better define the cellular behaviors exhibited by migrating murine cranial neural crest cells, we have analyzed a series of quantifiable migratory parameters, specifically focusing on migratory capacity and cell shape dynamics.  Repeat for other time-lapse files. NOTE: Trajectory plots can be used to visualize the directness of individual cell paths for a given cell condition or state over the course of time-lapse movies ( Figure 4A). Distance and speed data stored in the .txt files can then be used for further analysis. can be used to quantify how cell shape dynamics change over time (Figure 4C) or how morphology may be altered under different cell treatments.

Representative Results
Using the procedure demonstrated here, mouse embryos were dissected from the uterus, and extraembryonic tissues were removed ( Figure  1A-D). Embryos were somite staged (using only embryos at 5-8 somites (ss), Figure 1E,F). The cranial neural plate was then dissected and the neuroepithelium was isolated. Mesodermal cells, identified as loose, circular, mesenchymal cells, were gently brushed off (Figure 1G-L). The anterior neural plate can be explanted whole, in which case the neural crest tissue will emerge laterally and expand radially around the explant, or each neural plate border (right and left) can be explanted separately. This is particularly useful when explanting from genetic mutants.
Within 24 h, a region of premigratory (epithelial) cranial neural crest can clearly be seen surrounding the neural plate explant ( Figure 2B). Furthermore, a subpopulation of neural crest cells have undergone epithelial to mesenchymal transition and appear fully mesenchymal ( Figure  2). Thus, we have several concentric rings of distinct cells, with the neural plate (NP) in the center, the premigratory neural crest (pNC) in the intermediate circle, and a population of migratory neural crest (mNC) in the outside ring ( Figure 2B). In order to trace NC cells, it is possible to use genetically modified mouse models as we show in Figure 2C. In this case, we have used the neural crest specific Wnt1::Cre;Rosa mTmG which results in NC cells being labeled in green. In these mice, cells express membrane tomato (mT, in red) unless they are expressing Cre recombinase. Recombination leads to cells expressing membrane green fluorescent protein (GFP, in green). The red cells shown at the center of the explant are neural plate cells. Some dorsal neural plate cells also express GFP; for long term culture, we would excise all of the cells in the center. For our purposes, the purity of the explant is sufficient to track the different neural crest cell populations. Where higher purity of the neural crest is necessary, this genetic labeling strategy can be combined with fluorescent activated cell sorting (FACS) to ensure purity of the population. Alternatively, it is possible to fix the explants and identify the NC population with antibody labeling.
It was also evident by 24 h that the characteristic concentric rings of premigratory and fully migratory NC cells of the explant cultures was not dependent nor governed by matrix choice (Figure 3). Explant cultures plated on both an ECM-based hydrogel and fibronectin formed comparable explant structures, comprising the three cell populations, NP, pNC and mNC (Figure 3A,C). Neural crest cell morphology was also comparable between those plated on the ECM-based hydrogel and fibronectin (Figure 3B,D). However, explants plated on fibronectin produced cells with more prominent lamellipodia at the cell leading edge, seemingly more polarized in the direction of migration (Figure 3B,D).

Discussion
Studying mammalian neural crest cells has been a challenge for scientists because of the in utero nature of mammalian development. In vivo studies are difficult to set up, as the embryo must be manipulated under conditions that mimic life in the uterus. In practice, it is nearly impossible to reproducibly culture these (E8+) embryos for longer than 24 h, especially for live imaging. Furthermore, neural crest induction and migration occur concurrently with neural tube closure and embryonic turning in the mouse; this is a crucial and stressful morphogenetic event, which frequently fails when embryos are cultured ex utero. Thus, the success rate of ex utero approaches is generally low. The use of immortalized NC cells 21 is a useful tool to reduce animal use and it may provide a better source of neural crest cells for long-term analysis, transfection, and enrichment studies. However, there is clearly a need to reliably culture primary neural crest cells. Our method is applicable to mouse knockout or conditional genetic models. A comparable method to ours has been described for other neural crest populations 20 ; however, our method thoroughly describes the step-by-step isolation of murine cranial NC cells. We also describe the use of different matrices as well as the migration analysis procedure in detail.
To achieve consistent results, we found that special attention paid towards staging during the selection of embryos. Not surprisingly, the number of somites correlates with different stages in cranial NC development. Therefore, the knowledge of the embryo anatomy is very important before acquiring any experimental data. This approach can then be adapted towards isolating discrete populations of neural crest cells, depending on the biological question and the target cells.
Once the embryos are selected and dissected, mesodermal cells can be readily distinguished and should be removed to allow better visualization and to reduce contamination. For longer-term cultures, the neural plate tissue can be removed at 24 h of plating in order to prevent contamination by neural tissues. A further refinement could be the use of fluorescent lineage labeling (for example, using a Wnt1::cre or Sox10::creERT drivers combined with fluorescent reporters 24,25 ) to distinguish neural crest cells from other tissues as shown in Figure 2C.
Previous reports have highlighted the potential of plating mouse NC explant cultures on different matrices, most commonly on commercial ECM hydrogels, fibronectin and collagen I 20,21,26 . In our hands, mouse cranial NC explant cultures are successfully grown on all three matrices, at concentrations specified in original reports (data not shown). The initial refined approach we adapted for our NC explant cultures used a commercial hydrogel as the matrix of choice, which primarily consisted of laminin and collagens 21 (Figure 3A-B). However, composition of this hydrogel is not clearly defined, with unknown growth factor and protein content. As such, we have since shifted our approach to plating mouse NC explant cultures on fibronectin (Figure 3C-D). Fibronectin is well defined and highly expressed in the ECM and basement membranes along which NC cells migrate in vivo 28,29,30 . To optimize a fibronectin matrix that best replicates neural crest cell migration and morphology as seen using the hydrogel, we compared NC cell behaviors exhibited on the hydrogel against a titration of 0.25-30 µg/mL fibronectin, and defined 1 µg/mL fibronectin as providing ideal properties (data not shown). We believe that this preliminary work may help establish a framework for the systematic comparison of matrices, such as fibronectin, against those previously described, namely collagen and laminin 32,33,34 . It would be especially interesting to compare mouse NC cell migratory capacity on fibronectin versus collagen I, given that collagen-IA1 is endogenously secreted by mouse, avian and human NC cells 28,30,31,32 . Collagen I is therefore as relevant as fibronectin in the consideration of matrix choice. It is also worth acknowledging that the bioavailability of growth factors in the media may be altered by different matrix components, especially given the high serum content of our media. To overcome this, we are currently working to produce serum-free defined culture conditions. These defined media are successfully used in neural crest induction protocols in the pluripotent stem cell field, but require further optimization for our NC explant culture system 33,34 . Our work may also serve as a starting point for refining conditions for other types of NC cells such as cardiac and trunk NC, and for subsequent studies of NC differentiation. Most importantly, this protocol allows the isolation of cranial NC cells for a variety of applications. We envision studies on directed migration, 3-D migration and invasion. Cells isolated in this manner can be treated in vitro for a number of analyses. For example, cells can readily be treated using different small molecules to target specific proteins, they can be treated at defined time points, and washout experiments can be designed to determine recovery of cell behaviors (Figure 4). Longer term culture for transfection and differentiation assays is possible, as well as passage of cells (data not shown). However, the viability, the capacity of cell renewal and multipotency should be validated after passaging. Cells plated on glass coverslips can also be used in immunofluorescent staining protocols, following live imaging. Finally, this approach represents a tremendously powerful system for studying migration of NC from genetic mouse models 22,23,24,25 .

Disclosures
The authors have no conflicts of interest.