Reinvestigation of the role of the optic vesicle in embryonic lens induction

The induction of the lens by the optic vesicle in amphibians is often cited as support for the view that a single inductive event can lead to determination in a multipotent tissue. This conclusion is based on transplantation experiments whose results indicate that many regions of embryonic ectoderm which would normally form epidermis can form a lens if brought into contact with the optic vesicle. Although additional evidence argues that during normal development other tissues, acting before the optic vesicle, also contribute to lens induction, it is still widely held, on the basis of these transplantation experiments, that the optic vesicle alone can elicit lens formation in ectoderm. While testing this conclusion by transplanting optic vesicles beneath ventral ectoderm in Xenopus laevis embryos, it became apparent that contamination of optic vesicles by presumptive lens ectoderm cells can generate lenses in these experiments, illustrating the need for adequate host and donor marking procedures. Since previous studies rarely used host and donor marking, it was not clear whether they actually demonstrated that the optic vesicle can induce lenses. Using careful host and donor marking procedures with horseradish peroxidase as a lineage tracer, we show that the optic vesicle cannot stimulate lens formation in neurula- or gastrula-stage ectoderm of Xenopus laevis. Since the general conclusion that the optic vesicle is sufficient for lens induction rests on studies in many organisms, we felt it was important to begin to test this conclusion in other amphibians as well. Similar experiments were therefore performed with Rana Palustris embryos, since it was in this organism that optic vesicle transplant studies had originally argued that this tissue alone can cause lens induction. Under conditions similar to those used in the original report, but with careful controls to assess the origin of lenses in transplants, we found that the optic vesicle alone cannot elicit lens formation. Our data lead us to propose that the optic vesicle in amphibians is not generally sufficient for lens induction. Instead, we argue that lens induction occurs by a multistep process in which an essential phase in lens determination occurs as a result of inductive interactions preceding contact of ectoderm with the optic vesicle.     

Inductive interactions in the spatial and temporal restriction of lens-forming potential in embryonic ectoderm of Xenopus laevis

The process of lens cell determination in amphibians is currently viewed as one involving a series of inductive interactions. On the basis of previous investigations, these interactions are thought to begin during gastrulation when the presumptive foregut endoderm and then the heart mesoderm come into contact with the presumptive lens ectoderm. This earlier period of induction is followed by the later interaction of the optic vesicle with the lens-forming ectoderm. Transplantation experiments were performed to determine the relative significance of the early and later periods of induction in the process of lens cell determination in the anuran Xenopus laevis. Various ectodermal tissues were transplanted either into the lens-forming region of open neural plate stage host embryos or over the newly formed optic vesicle of later neurula stage embryos. All transplanted tissues were labeled with the intracellular marker horseradish peroxidase to assess the exact origins of any induced lens structures. The results indicate that all nonneural ectodermal tissues have some lens-forming potential early during gastrulation; however, this potential is restricted to the lens-forming region, and perhaps nearby regions, later in development during the time of neurulation. Furthermore, the results show that the optic vesicle is not a substantial inductor of the lens in tissues that have not been previously exposed to the earlier series of inductive interactions that take place during gastrulation and neurulation. Since the optic vesicle does not appear to be a sufficient inductor of the lens, these earlier inductive interactions are, therefore, essential in the process of lens cell determination in Xenopus. These earlier inductive interactions lead to a steady increase in what may be called a lens-forming bias in the presumptive lens ectoderm during this period of development. The eventual loss in the ability of nonlens ventral ectoderm to respond to these lens inductors is presumably the result of other determinative processes that occur in this tissue.     

Early tissue interactions leading to embryonic lens formation in Xenopus laevis

Our previous research has demonstrated that lens induction in Xenopus laevis requires inductive interactions prior to contact with the optic vesicle, which classically had been thought to be the major lens inductor. The importance of these early interactions has been verified by demonstrating that lens ectoderm is specified by the time it comes into contact with the optic vesicle. It has been argued that the tissues which underlie the presumptive lens ectoderm during gastrulation and neurulation, dorsolateral endoderm and mesoderm, are the primary early inductors. We show here, however, that these tissues alone cannot elicit lens formation in Xenopus ectoderm. Evidence is presented that presumptive anterior neural plate tissue (which includes the early eye rudiment) is an essential early lens inductor in Xenopus. The presence of dorsolateral mesoderm appears to enhance this response. These findings support a model in which an essential inductive signal passes through the plane of ectoderm during gastrula and early neurula stages from presumptive anterior neural tissue to the presumptive lens ectoderm. Since there is evidence for such interactions within a tissue layer in mesodermal and neural induction as well, this may be a general feature of the initial stages of determination of many tissues.     

Reliability of delta-crystallin as a marker for studies of chick lens induction

Induction of a lens by the optic vesicle of the brain was the first demonstration of how tissue interactions could influence cell fate during development. However, recent work with amphibians has shown that the optic vesicle is not the primary inducer of lens formation. Rather, an earlier interaction between anterior neural plate and presumptive lens ectoderm appears to direct lens formation. One problem with many early experiments was the absence of an unambiguous assay for lens formation. Before being able to test whether the revised model of lens induction applies to chicken embryos, we examined the suitability of using delta-crystallin as a marker of lens formation. Although delta-crystallin is the major protein synthesized in the chick lens, one or both of the two delta-crystallin genes found in chickens is transcribed in many non-lens tissues as well. In studies of lens formation where appearance of the delta-crystallin protein is used as a positive assay, synthesis of delta-crystallin outside of the lens could make experiments difficult to interpret. Therefore, polyacrylamide gel electrophoresis, immunoblotting, and immunofluorescence were used to determine whether the delta-crystallin messenger RNA detected in non-lens tissues is translated into protein, as it is in the lens. On Coomassie-blue-stained gels of several tissues from stage-22 embryos, a prominent protein was observed that co-migrated with delta-crystallin. However, on immunoblots, none of the non-lens tissues tested contained detectable levels of delta-crystallin at this stage. By imunofluorescence, delta-crystallin was observed in Rathke's pouch and in a large area of oral ectoderm near Rathke's pouch, yet none of the cells in these non-lens tissues showed the typical elongated morphology of lens fiber cells. When presumptive lens ectoderm or other regions of ectoderm from stage-10 embryos were cultured and tested for lens differentiation, both cell elongation and delta-crystallin synthesis were observed, or neither were observed. The results suggest that delta-crystallin synthesis and cell elongation together serve as useful criteria for assessing a positive lens response.     

A novel fork head gene mediates early steps during Xenopus lens formation.

Xlens1 is a novel Xenopus member of the fork head gene family, named for its nearly restricted expression in the anterior ectodermal placode, presumptive lens ectoderm (PLE), and anterior epithelium of the differentiated lens. The temporal and spatial restriction of its expression suggests that: (1) Xlens1 is transcribed initially at neural plate stages in response to putative signals from the anterior neural plate that transform lens-competent ectoderm to lens-biased ectoderm; (2) further steps in the process of lens-forming bias restrict Xlens1 expression to the presumptive lens ectoderm (PLE) during later neural plate stages; (3) interactions with the optic vesicle maintain Xlens1 expression in the lens placode; and (4) Xlens1 expression is downregulated as committed lens cells undergo terminal differentiation. Induction assays demonstrate that pax6 induces Xlens1 expression, but unlike pax6, Xlens1 cannot induce the expression of the lens differentiation marker beta-crystallin. In the whole embryo, overexpression of Xlens1 in the lens ectoderm causes it to thicken and maintain gene expression characteristics of the PLE. Also, this overexpression suppresses differentiation in the lens ectoderm, suggesting that Xlens1 functions to maintain specified lens ectoderm in an undifferentiated state. Misexpression of Xlens1 in other regions causes hypertrophy of restricted tissues but only occasionally leads ectopic sites of gamma-crystallin protein expression in select anterior head regions. These results indicate that Xlens1 expression alone does not specify lens ectoderm. Lens specification and differentiation likely depends on a combination of other gene products and an appropriate level of Xlens1 activity.     

Induction of the Eye Lens

In general terms embryonic induction involves the association of embryonic tissues and leads to tissue differentiation. It is one of the known essential processes leading to the normal development of embryos. However, despite its importance, very little is known about the mechanisms of inductive interactions. For example, what is the nature of communication between tissues, how does this communication effect the synthetic activity of the cells, and once a new pattern of synthesis has been established how is the sequence of events leading to tissue differentiation co-ordinated? The answers to these questions will come only from the intensive study of inductive interactions and tissue differentiation at all levels from the morphological to the molecular.
One of the best known examples of induction, at least superficially, is the differentiation of lens from head ectoderm after its interaction with optic vesicle. The popularity of this tissue with embryologists may be attributed to its accessibility of manipulation because of its position on the outside of the embryo. In addition, its distinct morphology and specific biochemical composition make it relatively easy to determine whether the lens differentiates after experimental treatment. About the turn of this century lens differentiation was thought to depend on the specific interaction of just two embryonic tissues, head ectoderm and neuro-ectoderm (optic vesicle). However, experimental analysis since then has revealed that this oversimplified view of lens induction is incorrect. In fact there is evidence that a large number of other tissues besides embryonic head ectoderm can differentiate into lens and that other conditions besides the presence of optic vesicle can induce lens differentiation. The purpose of this work is to review the evidence on lens induction and based on this, to determine what we know about the mechanism(s) controlling this process.     

FGF-mediated induction of ciliary body tissue in the chick eye

Upon morphogenesis, the simple neuroepithelium of the optic vesicle gives rise to four basic tissues in the vertebrate optic cup: pigmented epithelium, sensory neural retina, secretory ciliary body and muscular iris. Pigmented epithelium and neural retina are established through interactions with specific environments and signals: periocular mesenchyme/BMP specifies pigmented epithelium and surface ectoderm/FGF specifies neural retina. The anterior portions (iris and ciliary body) are specified through interactions with lens although the molecular mechanisms of induction have not been deciphered. As lens is a source of FGF, we examined whether this factor was involved in inducing ciliary body. We forced the pigmented epithelium of the embryonic chick eye to express FGF4. Infected cells and their immediate neighbors were transformed into neural retina. At a distance from the FGF signal, the tissue transitioned back into pigmented epithelium. Ciliary body tissue was found in the transitioning zone. The ectopic ciliary body was never in contact with the lens tissue. In order to assess the contribution of the lens on the specification of normal ciliary body, we created optic cups in which the lens had been removed while still pre-lens ectoderm. Ciliary body tissue was identified in the anterior portion of lens-less optic cups. We propose that the ciliary body may be specified at optic vesicle stages, at the same developmental stage when the neural retina and pigmented epithelium are specified and we present a model as to how this could be accomplished through overlapping BMP and FGF signals.     

Establishment of the scleral cartilage in the chick.

This study was undertaken to investigate the establishment of the scleral cartilage in the chick embryo. Johnston et al. (1974) has demonstrated that most of the cells of the scleral cartilage originate in the cranial neural crest. By means of a series of chorioallantoic grafts of pigmented retina, and its adherent periocular mesenchyme from stage 11 to 25, the present experiments show that the cranial neural crest cells arrive at the eye in sufficient numbers to form cartilage by stage 14. Pigmented retina, denuded of mesenchyme, from stage 16 embryos implanted into the head of stage 13 embryos induces cartilage formation in head mesenchyme. However, neither pigmented retina nor spinal cord could induce cartilage formation in chorioallantoic mesenchyme. Combination grafts of cranial neural crest and presumptive optic vesicle developed neural tissue, pigmented retina, and in some cases sclera-like cartilage. Thus, periorbital mesenchyme, derived largely from cranial neural crest, at about stage 14 develops the scleral cartilage in response to induction by the pigmented retina.     

The pattern of protein and glycoprotein synthesis in presumptive lens and non-lens ectoderm of the chicken embryo

Summary Lens induction is a classic example of the tissue interactions that lead to cell specialization during early vertebrate development. Previous studies have shown that a large region of head ectoderm, but not trunk ectoderm, of 36 h (stage 10) chicken embryos retains the potential to form lenses and synthesize the protein δ-crystallin under some conditions. We have used polyacrylamide gel electrophoresis and fluorography to examine protein and glycoprotein synthesis in presumptive lens ectoderm and presumptive dorsal (trunk) epidermis to look for differentiation markers for these two regions prior to the appearance of δ-crystallin at 50 h. Although nearly all of the proteins incorporating³H-leucine were shared by presumptive lens ectoderm and trunk ectoderm, these two regions showed more dramatic differences in the incorporation of³H-sugars into glycoproteins. when non-lens head ectoderm that has a capacity for lens formation in vitro was labeled, a hybrid pattern of glycoprotein synthesis was discovered: glycoproteins found in either presumptive lens ectoderm or trunk ectoderm were oftentimes also found in other head ectoderm. Therefore, molecular markers have been identified for three regions of ectoderm committed to different fates (lens and skin), well before features of terminal differentiation begin to appear in the lens.     

Early Determination of Developmental Fate in Presumptive Intestinal Endoderm of the Chicken Embryo

The endodermal epithelia of esophagus, proventriculus and gizzard of 6-day chicken embryos can form glands and express embryonic chicken pepsinogen (ECPg), when they are subjected to the influence of proventricular mesenchyme, while intestinal epithelium of the same age cannot respond to the inductive influence of proventricular mesenchyme. We attempted in this paper to know whether this regional difference of epithelia to respond to mesenchymal influence originates very early in development or it is established gradually in the course of development of digestive tract.
The young presumptive intestinal endoderm taken from embryos having 15–20 somites was associated and cultivated with 6-day proventricular mesenchyme. The presumptive intestinal endoderm never expressed ECPg although it formed gland-like structures. In the control explants composed of presumptive stomach endoderm and proventricular mesenchyme, glands were formed and gland cells expressed ECPg detected by immunocytochemistry and in situ hybridization.
These results indicate that the developmental fate of presumptive intestinal endoderm is determined rather strictly at very early developmental stage, and suggest that the segregation of at least two cell lineages occurs early in the development; one which can express ECPg under the influence of proventricular mesenchyme, and another one which cannot express ECPg and differentiates mainly into intestinal epithelium.     

Extraocular dorsal signal affects the developmental fate of the optic vesicle and patterns the optic neuroepithelium

Dorsal-ventral (DV) specification in the early optic vesicle plays a crucial role in the proper development of the eye. To address the questions of how DV specification is determined and how it affects fate determination of the optic vesicle, isolated optic vesicles were cultured either in vitro or in ovo. The dorsal and ventral halves of the optic vesicle were fated to develop into retinal pigment epithelium (RPE) and neural retina, respectively, when they were separated from each other and cultured. In optic vesicles treated with collagenase to remove the surrounding tissues, the neuroepithelium gave rise to cRax expression but not Mitf, suggesting that surrounding tissues are necessary for RPE specification. This was also confirmed in in ovo explant cultures. Combination cultures of collagenase-treated optic vesicles with either the dorsal or ventral part of the head indicated that head-derived factors have an important role in the fate determination of the optic vesicle: in the optic vesicles co-cultured with the dorsal part of the head Mitf expression was induced in the neuroepithelium, while the ventral head portion did not have this effect. The dorsal head also suppressed Pax2 expression in the optic vesicle. These observations indicate that factors from the dorsal head portion have important roles in the establishment of DV polarity within the optic vesicle, which in turn induces the patterning and differentiation of the neural retina and pigment epithelium.     

Transfilter Lens Induction in Avian Embryo

The directive influence of the optic vesicle during lens induction of 2-day chick embryos was studied in vitro. Trunk ectoderm was chosen for the responding tissue. This uncommitted ectoderm formed distinct lentoid bodies when grown together with the optic vesicle. Crystallin synthesis was demonstrated in the lentoids with fluorescein labelled antiserum.
By interposing filters of different thicknesses and pore sizes between the interacting tissues, three questions were put forward: (1) Whether the intimate contact between the interactants was essential for induction, (2) how far the inductive influence of the optic vesicle extended, and (3) what was the smallest pore through which the inductive influence could penetrate.
The inductive influence reached across a Millipore filter with a thickness of 100 μm. It also penetrated a dialyzer membrane, allowing passage of molecules with molecular weights (MW) of less than 12,000 daltons. It was concluded that the directive, inductive signal(s) passing from the optic vesicle, diffused in the extracellular space as far as 100 μm and had a molecular weight of less than 12,000 daltons.     

Distinct Requirements for Cranial Ectoderm and Mesenchyme-Derived Wnts in Specification and Differentiation of Osteoblast and Dermal Progenitors

The cranial bones and dermis differentiate from mesenchyme beneath the surface ectoderm. Fate selection in cranial mesenchyme requires the canonical Wnt effector molecule β-catenin, but the relative contribution of Wnt ligand sources in this process remains unknown. Here we show Wnt ligands are expressed in cranial surface ectoderm and underlying supraorbital mesenchyme during dermal and osteoblast fate selection. Using conditional genetics, we eliminate secretion of all Wnt ligands from cranial surface ectoderm or undifferentiated mesenchyme, to uncover distinct roles for ectoderm- and mesenchyme-derived Wnts. Ectoderm Wnt ligands induce osteoblast and dermal fibroblast progenitor specification while initiating expression of a subset of mesenchymal Wnts. Mesenchyme Wnt ligands are subsequently essential during differentiation of dermal and osteoblast progenitors. Finally, ectoderm-derived Wnt ligands provide an inductive cue to the cranial mesenchyme for the fate selection of dermal fibroblast and osteoblast lineages. Thus two sources of Wnt ligands perform distinct functions during osteoblast and dermal fibroblast formation.