Induction and specification of cranial placodes.
Cranial placodes are specialized regions of the ectoderm, which give rise to various sensory ganglia and contribute to the pituitary gland and sensory organs of the vertebrate head. They include the adenohypophyseal, olfactory, lens, trigeminal, and profundal placodes, a series of epibranchial placodes, an otic placode, and a series of lateral line placodes. After a long period of neglect, recent years have seen a resurgence of interest in placode induction and specification. There is increasing evidence that all placodes despite their different developmental fates originate from a common panplacodal primordium around the neural plate. This common primordium is defined by the expression of transcription factors of the Six1/2, Six4/5, and Eya families, which later continue to be expressed in all placodes and appear to promote generic placodal properties such as proliferation, the capacity for morphogenetic movements, and neuronal differentiation. A large number of other transcription factors are expressed in subdomains of the panplacodal primordium and appear to contribute to the specification of particular subsets of placodes. This review first provides a brief overview of different cranial placodes and then synthesizes evidence for the common origin of all placodes from a panplacodal primordium. The role of various transcription factors for the development of the different placodes is addressed next, and it is discussed how individual placodes may be specified and compartmentalized within the panplacodal primordium. Finally, tissues and signals involved in placode induction are summarized with a special focus on induction of the panplacodal primordium itself (generic placode induction) and its relation to neural induction and neural crest induction. Integrating current data, new models of generic placode induction and of combinatorial placode specification are presented.
PubMed ID: 16677629
Article link: Dev Biol.
Genes referenced: dlx3 eya1 msx1 pax2 pax3 pax6 pax8 six1 six4
Article Images: [+] show captions
|Fig. 1. Cranial placodes in vertebrates. (A) Cranial placodes in a 10- to 13-somite-stage chick embryo (modified from Streit, 2004; based on D'Amico-Martel and Noden, 1983; Bhattacharyya et al., 2004). In amniotes, profundal and trigeminal placode are commonly referred to as ophthalmic and maxillomandibular placode of the trigeminal nerve, respectively. (B) Cranial placodes in a tailbud stage Xenopus embryo (modified from Schlosser and Northcutt, 2000). (C) Schematic summary of morphogenesis and cellular derivatives of various cranial placodes. Invagination occurs in adenohypophyseal, olfactory, lens, and otic placodes. Moreover, in all placodes except the lens placode, some cells migrate away from the placodal epithelium as mesenchymal cells to form sensory neurons, secretory cells, or glial cells. In lateral line placodes, another subset of cells migrates along the basement membrane and forms the lateral line primordia (modified from Schlosser, 2005).|
|Fig. 2. Fate maps of cranial placodes at late gastrula or neural plate stages. The boundary of the (prospective) neural plate ( here defined as the precursor of the entire neural tube, which includes the inner but not the outer neural folds) is indicated by the broken line. All placodes originate from teh pre-placodal region in the non-neural ectoderm within and immediately adjacent to the outer neural folds. (A) Fate map of late gastrula zebrafish embryos with substantial overlap between ectodermal regions giving rise to different placodes (modified from Kozlowski et al., 1997). (B) Fate map of neural plate stage chick embryo (0–1 somite stage) with extensive overlap between various prospective placodes (based on data from Streit, 2002; Bhattacharyya et al., 2004). (C) Fate map of neural plate stage salamander (Ambystoma) embryo (based on Carpenter, 1937). Recently, otic and lateral line placodes in Ambystoma were reported to originate from the outer neural folds themselves and, thus, from slightly more medial positions than depicted here (Northcutt, 1996). HN: Hensen's node.|
|Fig. 3. Six1, Six4, and Eya1 define a panplacodal primordium as illustrated for Xenopus. (A) Six1 is first expressed in a horseshoe-shaped domain around the neural plate. (B) At tailbud stages, Six1 expression continues in all placodes. The position of the fused profundal and trigeminal ganglia is indicated by asterisks (the respective placodes are already disappearing at this stage). (C) Position of the panplacodal primordium (red) relative to the neural plate (gray) and neural crest (blue). The gray broken line indicates the border of the neural plate proper (i.e. the entire ectoderm, which will become incorporated into the neural tube). Solid colored lines enclose areas of expression of various transcription factors, while the broken lines (Dlx3 and Msx1) enclose areas, from which expression is excluded. Note that the region of the outer neural folds immediately peripheral to the hatched gray line is occupied by neural crest cells laterally but by the panplacodal primordium anteriorly. Although no fate map is available, gene expression patterns suggest that placodes arise approximately from the following regions as indicated by white letters: the preoral ectoderm gives rise to the stomodeum with the adenohypophyseal placode and to the olfactory placodes; the lateral domain of Pax6 expression gives rise to lens and trigeminal placodes; the lateral Pax3 domain gives rise to the profundal placode; finally, the domain of Pax8 and Pax2 expression gives rise to the posterior placodal area, from which otic, lateral line, and epibranchial placodes develop. Expression data are based on the following sources: Bang et al. (1997), Bellefroid et al. (1998), Zygar et al. (1998), Heller and Brändli (1999), Feledy et al. (1999), Ermakova et al. (1999), Hollemann and Pieler (1999), Zuber et al. (1999), Zhou et al. (2000), Sasai et al. (2001), Hartley et al. (2001), Luo et al. (2001a), Schweickert et al. (2001), Schlosser and Ahrens (2004). Abbreviations: Ad: adenohypophyseal placode; EB: epibranchial placodes; L: lens placode; LL: lateral line placodes; Ol: olfactory placode; Ot: otic placode; Pr: profundal placode; V: trigeminal placode.|
|Fig. 4. Summary of placodal development in Xenopus laevis in lateral views (modified from Schlosser and Ahrens, 2004; stages after Nieuwkoop and Faber, 1967). (A, B) Gene expression domains (colored outlines) during neural plate (A) and neural fold stages (B). The panplacodal primordium (red) is apposed to the neural plate (gray) anteriorly and to neural crest (blue) laterally (see also Fig. 3C). Arrows in panel B indicate shifts of placodal expression domains due to neural fold elevation (dorsal shift) and wedge-like expansion of anterior ectoderm accompanying the formation of the optic vesicles. Green stars identify three areas of Pax6 expression that will contribute (from anterior to posterior) to adenohypophyseal and olfactory placode (light green), lens placode (blue green), and trigeminal placode (dark green). Six1 and Eya1 expressions are downregulated at neural fold stages in the regions of prospective lens placode and cement gland (black asterisk). (C–F) Development of placodes after neural tube closure until late tailbud stages. Drawings are based on reconstructions of ectodermal thickenings from serial sections (after Schlosser and Ahrens, 2004; Schlosser and Northcutt, 2000). The adenohypophyseal placode, which is located medial to the ventral part of the olfactory placode, is not shown in these lateral views. Various green colors identify placodes or prospective placodes expressing Pax6, yellow identifies the profundal placode expressing Pax3, and brown, pink, and orange jointly identify the posterior placodal area expressing Pax2 and Pax8, with pink being reserved for the subregion forming the otic placode/ vesicle and orange for the subregions forming the epibranchial placodes. At stage 21, the posterior placodal area is divided into an anterior and a posterior subregion of thickened ectoderm, separated ventrally by an indentation and dorsally by a region of thinner ectoderm (between broken black lines), while the prospective otic placode (pink) is identifiable as a particularly prominent thickening. The broken blue lines in panel C indicate neural crest streams as reconstructed from serial sections (Schlosser and Northcutt, 2000). The arrow in panel C indicates that the posterior placodal area expands posteriorly at early tailbud stages. Brown arrowheads in panels D–F indicate developing lateral line primordia. Abbreviations: Ad/Ol: anterior placodal area, from which adenohypophyseal (Ad) and olfactory placodes (Ol) develop; AV: anteroventral lateral line placode; cg: cement gland; Hp1: first hypobranchial placode; L: prospective lens placode (hatched outline), lens placode or lens (invagination of placode between stage 27 and 33/34); LL/Ot/EB: posterior placodal area, from which lateral line (LL), otic (Ot), and epibranchial (EB) placodes develop; M: middle lateral line placode; Ol: olfactory placode; Ot: otic placode or vesicle (invagination of placode between stage 24 and 33/34); P: posterior lateral line placode; Pr: profundal placode; V: trigeminal placode; VII: facial epibranchial placode; IX: glossopharyngeal epibranchial placode; X1: first vagal epibranchial placode; X2/3: second and third vagal epibranchial placodes (fused).|
|Fig. 6. Schematic summary of transcription factor expression domains in the placodal ectoderm of neural plate stage Xenopus embryos. Transcription factors that control neuronal determination and differentiation and are expressed in all neurogenic placodes but excluded from adenohypophyseal and lens placodes are not included. The position of various prospective placodes within the panplacodal primordium (faint red) is indicated by colored rectangles. Colored lines enclose expression domains of the transcription factors listed at neural plate stages. Listing of transcription factors without asterisk refers to their expression domains established at neural plate stages, while listing of transcription factors preceded by asterisks refers to expression domains established at later stages. Question marks indicate tentative assignments because precise domain boundaries cannot be determined unambiguously from published information. For simplicity, some details of expression are ignored (for example, Msx1 and Msx 2 are strongly expressed only in the dorsal part of the indicated region). For details and references, see Fig. 5. Note that there are two foci of transcription factor expression, one centered on an anterior placodal area (with prospective adenohypophyseal and olfactory placodes), the other centered on a posterior placodal area (with prospective lateral line, otic, and epibranchial placodes). However, the spatial extension of various expression domains differs resulting in two nested hierarchies, which are overlapping in the regions of profundal and trigeminal placodes. For detailed explanation, see text. Abbreviations: Ad: adenohypophyseal placode; EB: epibranchial placodes; L: lens placode; LL: lateral line placodes; Ol: olfactory placode; Ot: otic placode; Pr: profundal placode; V: trigeminal placode.|
|Fig. 7. Combinatorial multistep model for placode induction and specification. Expression of various transcription factor genes involved in placode specification (colored rectangles and triangles, A–J) may be regulated by particular combinations of upstream factors including inducing signaling molecules (I1–I8) as well as other transcription factors, which bind to their cis-regulatory region. Upstream factors may be activating (arrows; broken arrows are used for inducing signals because these act indirectly via transcription factors downstream of signaling cascades) or repressing (bars). In the hypothetical case illustrated, upstream factors are assumed to be singly necessary but only jointly sufficient for activation of the downstream gene. The expression domain of a transcription factor (enclosed by colored lines; the red area represents the panplacodal primordium) is determined by the spatial extent of the expression domains of inducers and transcription factors acting upstream (indicated by black and gray lines for inducers and colored ellipses for transcription factors). The expression of transcription factors specific for individual placodes (lower panel) depends on panplacodal (black), multiplacodal (dark gray), and placode-specific (light gray) inducers, which may act either directly or indirectly (i.e. mediated by other transcription factors with a panplacodal, multiplacodal, or placode-specific distribution). Some transcription factors (as illustrated for D), which show placode-specific expression at later stages, may initially be more broadly expressed but then become restricted because they require input from more localized inducers for sustained activation. This model implies that different placodes share some but not all of the inducers and transcription factors involved in their specification. Moreover, it proposes that at early stages of placode specification (upper panel) there are nested and partially overlapping regions differentially biased for development of different sets of placodes. Multiplacodal bias is introduced by transcription factors (such as B–D) which cover multiple prospective placodes and activate various placode-specific transcription factors, without being by themselves sufficient for their activation. Panplacodally expressed transcription factors such as A may, in addition, activate genes involved in the regulation of generic placodal processes such as proliferation, cell shape changes, and neurogenesis (not shown).|
|ig. 8. Establishment of sharp boundaries between differentially specified regions by switch-like cell fate decisions or compartmentalization. (A) Transcription factors A and B, which are involved in specification of two different domains (e.g. neural crest and placodes or two different individual placodes), may be induced by two different inducers (IA and IB, respectively) and have initially overlapping expression domains followed by sharpening of the boundary. (B) Overlapping expression domains as depicted in panel A may be due to coexpression of A and B proteins (red and blue dots) in individual cells (circles). Sharpening of the boundary in this case involves the differential upregulation of one but downregulation of the other transcription factor in individual cells (switch-like cell fate decisions), for example, due to their direct or indirect mutual transcriptional repression (lower panel). (C) Alternatively or additionally, overlapping expression domains as depicted in panel A may be due to mixing of cells expressing different transcription factors. Sharpening of the boundary may then involve the sorting of cells by non-random cell movements (arrows) (compartmentalization), for example mediated by mutually repelling cell adhesion molecules (CA, CB), activated by the different transcription factors (lower panel). (D) However, occurrence of cell movements per se does not necessarily imply cell sorting as depicted in panel C since random cell movements (arrows) may also occur before ectodermal areas are differentially specified to express different transcription factors.|
|Fig. 9. Models for generic placode induction. (A) Four models have been proposed to explain the positioning of the neural crest, panplacodal primordium, and epidermis around the anterior neural plate. All models share the assumption that in a first step neural induction occurs by a signal (turquoise arrows) from the axial mesoderm (dark gray circle). The Delay model suggests that the slowly spreading neural inducer then reaches more lateral ectoderm after a time delay, during which ectodermal competence has changed in an autonomous fashion (more yellowish color). The Gradient model assumes that neural induction establishes a gradient of a morphogen (e.g., BMP; orange) followed by the specification of different ectodermal fates at different threshold concentrations. The Neural plate border state model proposes that a special neural plate border region (pink) is induced first followed by the induction of neural crest and panplacodal primordium by additional signals from adjacent ectoderm and mesoderm (blue and red arrows). Different versions of this model suggest (from left to right) that the neural plate border is established (1) by a morphogen gradient (orange), (2) by epidermal–neural interactions (pink arrows), or (3) by initial induction of a border state throughout the dorsal ectoderm followed by induction of a proper neural plate in the center (green arrows). The Binary competence model suggests that neural induction establishes a dorsal ectodermal competence region (possibly at a particular threshold of a BMP gradient), which maintains competence (light green) to form neural plate and neural crest and has a neural default fate, while ventral ectoderm cell-autonomously loses neural/neural crest competence but retains competence (light yellow) to adopt epidermal or placodal fates and has an epidermal default fate. Subsequently, additional signals from adjacent ectoderm and mesoderm induce neural crest and panplacodal primordium (blue and red arrows). (B) Summary of tissues and signals involved in induction of neural crest (left side; blue arrows and bars) and panplacodal primordium (right side, red arrows and bars) from the perspective of the Binary competence model. In addition to the cranial neural plate, its continuation into upper trunk levels is depicted. Whereas signals from prospective epidermis, neural plate border region, and paraxial mesoderm including Wnts, FGFs, and BMPs induce neural crest at the border of the neural/neural crest competence region, FGFs together with BMP and Wnt antagonists from the anterior neural plate and the cranial dorsolateral endomesoderm induce the panplacodal primordium at the border of the epidermal/placodal competence region. Because neural crest is specified earlier than the panplacodal primordium, it may escape the action of BMP and Wnt antagonists from the neural plate. Note that in this model differential ectodermal competence of the dorsal and ventral ectodermal territories defines the ventral limit of neural crest induction and the dorsal limit of generic placode induction. The ventral and posterior limits of generic placode induction are probably imposed by the extent of the signaling center producing BMP and Wnt antagonists in the cranial dorsolateral endomesoderm, which protects the overlying ectoderm from BMP and Wnt signals only in the dorsolateral head region. In contrast, the dorsal limits of neural crest induction may be set by thresholds of availability of the signals (Wnts, BMP, FGF) required for neural crest induction, while the anterior limits of crest induction are probably imposed by an unknown repressing signal that is rostrally confined.|