Most studies on generation of neurons in vertebrates have focused on the central nervous system, which is composed of the brain and the spinal cord. The peripheral nervous system contains sensory neurons whose neurogenesis has been much less explored. In the head, ectodermal placodes contribute to a majority of the peripheral neurons, including sensory neurons in the cranial ganglia of the trigeminal V, facial VII, glossopharyngeal IX, and vagal X nerves responsible for somatosensation, general visceral sensation, and gustation (reviewed in [1, 2]). They also give rise to essential components of the paired sense organs (lens, inner ear, and olfactory epithelium). Placodes are derived from discrete, usually thickened, regions of the embryonic head ectoderm in both sides of the forming neural tube. Placodes that form cranial ganglia are exclusively fated to become sensory neurons in the distal portions of the ganglia [3, 4]. The generation of placodal neurons from the embryonic head epithelium is an intriguing process that differs from neurogenesis in the central nervous system (CNS), which takes place entirely within the neural tube.
Ectodermal placodes together with another embryonic cell population, the neural crest, are responsible for formation of the entire peripheral nervous system in vertebrates [1, 3]. Compared with neural crest, however, much less is known about the mechanisms that govern formation of placodes. Several signaling pathways, including Wnts and Fgfs, have been implicated in playing a role in induction and differentiation of trigeminal (V) and epibranchial (forming the VII, IX, and X nerves) placodes [2, 5–7]. However, the dynamic cellular processes by which these placode-derived cells delaminate from the epithelial ectoderm, migrate and coalesce in the underlying mesenchyme to form sensory ganglia remain poorly understood.
Placode formation involves induction of placode-specific fates and often acquisition of thickened columnar morphology followed by two key steps: first, the birth of neuronal cells in the ectoderm and the second, their detachment and migration to the site of ganglion assembly. The first step likely involves intricately regulated patterns of mitoses to generate the proper number and position of placodal neurons in the ectoderm. Previous studies show that placodal cells differentiate as neurons in the ectoderm prior to their delamination, as evidenced by co-labeling of fluorescently tagged ingressing placodal cells by GFP or vital dye with pan-neuronal markers Islet-1, neuronal beta-III tubulin (TuJ1), and neurofilament [8, 9], and on cell morphology , showing that by the time of delamination placodal cells are already neurons. Furthermore, neuronal markers are detected in some scattered individual cells within the placodal ectoderm prior to placodal cell ingression [4, 8]. These findings suggest that neurogenesis takes place in the embryonic ectoderm, and that placodal cells delaminate as neuronal cell types. Whether ectoderm cells divide symmetrically or asymmetrically to give rise to placode and non-placode cell types and where they divide along the apical-basal sides of the ectoderm remains largely unknown. Such differential modes of division (symmetric versus asymmetric) have been implicated in cell fate decisions in the CNS [11, 12]. Phospho-histone H3 staining suggests that mitosis occurs apically in epibranchial placodes, and sectioned tissue suggests that placodes are pseudostratified in the embryonic epithelium ; however, this has yet to be analyzed directly by live-cell imaging.
The second step requires that placodal neurons make a transition from an epithelial to a mesenchymal like state in order to detach from the ectoderm. This process is often referred to as either ingression or delamination. However, placodal cells appear to undergo a different process of delamination than that of neural crest that undergo an epithelial-to-mesenchymal transition (EMT). For example, placode cells do not express typical EMT genes (i.e. Snail2 and the GTPase RhoB [13, 14]) and express neuronal markers at the time of delamination [8, 10]. How placodal cells change morphology to detach and acquire motility and how they migrate from the epithelial ectoderm remains elusive. Therefore, analyzing these cells in real time promises to reveal new insights into placodal cell behavior during development.
To date, in vivo imaging of placodal cell ingression in whole intact embryos remains optically difficult, largely due to the insufficient z-axis resolution in three-dimensional fluorescence microscopy. In particular, poor cellular resolution along the z-axis prevents clear delineation of cells in the surface epithelium from those that have migrated deep inside the head. To circumvent these issues, we have developed an imaging assay using cranial slice culture to monitor placodal cell behavior in real time at a single cell resolution. This novel imaging assay is based on adaptation of a long-term embryo slice culture imaging system previously established for chick spinal cord . Similar slice imaging assays have been powerful in elucidating the dynamics of migrating and dividing neuronal cells in the mammalian cortex [15–17]. Here, we show that this imaging assay is highly effective for capturing different events in placode formation, ranging from cell proliferation, delamination and migration to assembly of these cells into ganglia. We focus on trigeminal placode formation as these form the largest of the cranial sensory ganglia, and are the first to form, making them highly tractable for imaging early placode formation.