|
|
||||||||
1From the Department of Ophthalmology and Visual Science and Hermann Eye Center, University of Texas Medical School at Houston, Houston, Texas; and the 2Department of Ophthalmology, Harvard Medical School, Boston, Massachusetts.
| Abstract |
|---|
|
|
|---|
METHODS. Zebrafish embryos and larvae at various stages of development were used for documenting corneal morphogenesis and differentiation. Corneal samples were collected from embryos, larvae, and adult zebrafish for histologic and electron microscopy analysis. Expression patterns of corneal polypeptides were investigated by immunostaining of sections.
RESULTS. The zebrafish cornea develops rapidly during embryogenesis, so that its three major layers, the epithelium, the stroma, and the endothelium, are well formed by day 3 postfertilization. The subsequent steps of corneal differentiation, such as the thickening of the corneal stroma, proceed relatively slowly. Several polypeptides are highly enriched in the epithelium or the stroma of the larval and adult zebrafish cornea and are excellent markers of corneal differentiation.
CONCLUSIONS. Development and differentiation of the zebrafish cornea are easily accessible to analysis. Anatomic and ultrastructural characterization of the zebrafish cornea demonstrates many similarities to the human cornea and provides the basis for the use of the zebrafish model both to analyze the basic genetic mechanisms of corneal development and to study the causes of corneal disease.
The understanding of corneal differentiation and the mechanisms that maintain its structure and function is of major basic scientific and medical importance. Congenital corneal and anterior segment abnormalities are estimated to cause 1% to 3% of childhood blindness in the western world.2 3 Common developmental malformations of the cornea include agenesis (the total absence of the cornea), megalocornea (enlarged cornea), microcornea (small cornea), and cornea plana (corneal curvature changed).4 They may have both genetic and nongenetic causes. Mutations in the PAX6 gene, for example, are one of the best-investigated genetic causes of corneal defects in humans.5 In addition to developmental abnormalities, corneal dystrophies are another heterogeneous group of inherited corneal diseases that can affect the five layers of the cornea.6 Frequently, they feature the deposits of abnormal material (cellular debris or proteins) within the cornea, which result in the loss of its normal transparency. Another presumable corneal dystrophy, keratoconus, is characterized by a noninflammatory progressive thinning of the cornea and a loss of visual acuity.7 The understanding of how the cornea develops, differentiates, and ages at the cellular and molecular levels is of key importance for the prevention, diagnosis, and treatment of corneal disorders.
The zebrafish is an excellent animal model for the studies of eye development. One of the major advantages of the zebrafish is the ability to perform genetic screens for mutants of embryonic development (reviewed in Ref. 8 ). During the last decade, the phenotypes of numerous zebrafish eye mutants have been characterized in extensive detail. This has been facilitated by an impressive assortment of analytical tools, including numerous gene expression markers, cell-fate mapping approaches, and mosaic analysis.9 Of importance, chemical mutagenesis in zebrafish can now be efficiently followed by positional cloning, an approach that has greatly benefited from the zebrafish genome project.10 11 Mutagenesis approaches in zebrafish are complemented by reverse genetic methods, such as antisense morpholino knockdown of gene function.12 13 Because of its accessibility to observation and manipulation, both during early embryogenesis and throughout the life of the organism, the zebrafish cornea is very well suited for genetic analysis.
Despite extensive embryological and genetic analysis of the zebrafish retina, little is known about the zebrafish cornea and the genetic mechanisms of its differentiation. One of the goals of this study is to provide a comprehensive description of this organ. This type of analysis is essential in the conduct of genetic studies in zebrafish and subsequently in the use of the zebrafish eye as a model for investigating human corneal diseases. The genetic analysis of the cornea using the zebrafish model requires a thorough description of both its embryonic development and its structure in the adult organism. In this report, we use light and electron microscopy (EM), as well as immunohistochemical techniques to provide a detailed analysis of corneal development and differentiation in zebrafish.
| Methods |
|---|
|
|
|---|
Histology
Zebrafish larvae were raised to the desired age and processed for histologic analysis using JB4 embedding medium, as described previously.16 17 Sections were examined by microscope (Axioscope; Carl Zeiss, Thornwood, NY), and images were collected with a digital camera (AxioCam; Carl Zeiss Meditec). Image files were processed using image-adjusting software (Photoshop 7.0; Adobe Systems, Mountain View, CA).
Electron Microscopy
Zebrafish larvae and dissected eyes from adult fish were fixed overnight at 4°C in 2.5% glutaraldehyde and 2% formaldehyde in cacodylate buffer (0.1 M), containing 0.08 M CaCl2. After fixation, the larvae were washed twice in cacodylate buffer for 15 minutes each, postfixed in 2% osmium tetroxide in cacodylate buffer for 1.5 hours at room temperature, rinsed twice in cacodylate buffer for 10 minutes, rinsed once in water, and stained in 2% uranyl acetate in maleate buffer (Ernst F. Fullam, Schenectady, NY) for 20 minutes. Dehydration in graded ethanol series and embedding in Epon (Polysciences, Warrington, PA) were performed as described previously.16 18 Samples were dehydrated through a graded ethanol series (30%100% ethanol), equilibrated in propylene oxide, and embedded in araldite resin. Thin sections were stained with 2% aqueous uranyl acetate and Reynolds lead citrate. Analysis was performed using with one of two electron microscopes (CM-10; Phillips, Eindhoven, The Netherlands; and JEM 100CX; JEOL, Tokyo, Japan). To assure that the fixative solution penetrates into the internal layers of the cornea, for the SEM analysis of the corneal endothelium, we prepared whole eyes as well as dissected corneas.
Immunohistochemistry
Zebrafish embryos and dissected eyes from adult fish were fixed in 4% paraformaldehyde (wt/vol), washed in PBST (0.1% Tween in PBS), infiltrated with 30% sucrose (wt/vol) in PBST, and embedded in freezing medium (TBS medium; Triangle Biomedical Sciences, Inc., Durham, NC). Cryosections (14 mm thick) were air dried for 10 minutes, washed in PBST, and blocked for 1 hour with 10% donkey serum in PBST. Sections were incubated with primary antibodies in the blocking solution for 2 to 5 hours at room temperature, or overnight at 4°C. The following antibodies and dilutions were used: anti-keratan sulfate proteoglycan (clone EFG-11, 1:500; Chemicon International Inc, Temecula, CA), anti-keratin 3 (1:500; ICN Pharmaceuticals, Costa Mesa, CA), anti-BIGH3 (1:500; Santa Cruz Biotechnology, Santa Cruz, CA), and anti-collagen type I (1:500; Santa Cruz Biotechnology). After they were washed three times in PBST, the sections were stained with appropriate combinations of FITC-, Cy3-, or Cy5-conjugated secondary antibodies (1:500; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) for 2 hours, and washed two times in PBST. BODIPY FL phallacidin (1:40; Invitrogen-Molecular Probes, Carlsbad, CA) and 4',6-diamidino-2-phenylindole dihydrochloride (DAPI; 1:1000; Invitrogen-Molecular Probes) were used to label F-actin and nuclei, respectively. Sections were incubated with these reagents for 30 minutes, and mounted in medium consisting of 50% glycerol and 2% propyl gallate (Sigma-Aldrich Co., St. Louis, MO) in 0.2 M TrisHCl (pH 8.0). Staining results were analyzed using a confocal microscope (LSM Meta; Carl Zeiss). Human and macaque corneal tissue (obtained from the Lions Eye Bank of Texas and generously provided by Daniel Felleman of University of Texas Medical School at Houston, respectively) was sectioned at 20 to 30 µm and stained as just described.
| Results |
|---|
|
|
|---|
|
The zebrafish corneal stroma, located basal to the Bowmans layer and distal to the Descemets membrane, is a prominent structure, mostly composed of connective tissue. It is approximately 6 µm thick, contributes approximately 30% of the corneal thickness, and consists of many tightly packed strata of collagen bundles. Collagen fibers run parallel to each other within each stratum, and are arranged orthogonal to each other in adjacent layers (Figs 1A 1D) , providing structural rigidity. The stroma is populated by keratocytes, flat cells that extend their cytoplasmic processes radially in all directions (Figs 1A 1E) . Posterior to the stroma, the Descemets layer is secreted as a basement membrane of the corneal endothelium. It is approximately 0.15 µm thick at 6 mpf (Fig. 1F) .
The endothelium, the innermost component of the cornea, is a monolayer of polygonal cells. Its internal surface contains numerous membrane folds (Fig. 1G) , which we have observed after different fixation conditions (see Methods). The zebrafish endothelium is a distinct polymegathic monolayer that appears less firmly organized, compared to that in the human. We cannot entirely exclude the possibility that this difference is partly due to the fixation and SEM processing.
Rapid Embryonic Development of the Zebrafish Cornea
To document the early development of the zebrafish cornea, we collected embryos at stages ranging from 12 to 72 hours postfertilization (hpf) for histologic and electron microscopic analysis. During the first 3 days of embryonic development, rapid changes occur in the shape and structure of the cornea. At 12 hpf, the cornea and lens are not yet formed (Fig. 2A) . The presumptive corneal epithelium (epidermis) transforms into two layers of cells by 16 hpf (Fig. 2B) . During the next several hours, cells in the posterior layer appear to initiate lens formation, whereas the anterior layer appears to transform into the corneal primordium. The presumptive corneal epithelium (the surface cell layer) is continuous with the lens primordium at 18 hpf (Fig. 2C) . As the lens placode continues to differentiate during the next 6 hours (1824 hpf; Figs. 2D 2E 2F ), it gradually detaches from the overlaying epithelium. The separation of the corneal epithelium from the lens is largely completed by approximately 30 hpf (Fig. 2G) .
|
|
Maturation of Corneal Structure
The rapid growth of the eye is accompanied by a gradual maturation of the cornea. For the first 4 weeks of larval development, the corneal epithelium remains two-cell layers thick. During the same period, the stroma rapidly increases in thickness, due to the addition of collagenous lamellae by the epithelium, and the invasion of keratocytes. We counted the number of collagen fiber layers from four corneal samples at each stage of development examined. At 5 and 7 days postfertilization (dpf), 10 ± 0.82 and 12 ± 0.82 layers of fibers are seen in the stroma, respectively (Figs 4A 4B) . Subsequently, the stroma continues to thicken at the rate of approximately one fiber layer every 2 days, so that by 14 dpf it is approximately 18 ± 1.63 layers thick (Fig. 4C) . After that, the rate of stromal growth doubles to a single layer per day, and its thickness increases to approximately 32 ± 1.63 layers at 28 dpf (Fig. 4D) . The Bowmans layer is clearly formed by 5 dpf (Fig. 4A , asterisk), and significantly thickens between 14 and 28 dpf (compare Figs. 4C and 4D ). Another key corneal differentiation event is the appearance of keratocytes in the stroma. In zebrafish, keratocytes invade the stroma between 14 and 28 dpf (Fig. 4D) . At 28 dpf, the stroma is about one fifth of the total corneal thickness, whereas the epithelium comprises two thirds of corneal thickness.
|
Conserved Expression Patterns of Corneal Proteins in Vertebrates
To monitor the differentiation of the zebrafish cornea, we investigated the expression patterns of polypeptides found in corneas of other vertebrates: BIGH3, keratin 3 (K3), and corneal keratan sulfate proteoglycan (CKS). At 24 hpf, the BIGH3 protein is expressed at a low level, both in the retina and in the cornea, as revealed by staining with anti-BIGH antibody (Fig. 5A) . As the embryo develops, BIGH3 staining becomes more restricted to the cornea and the surface epidermis (Fig. 5B) , and in adult animals its expression is highly enriched in the corneal epithelium (Fig. 5C) , a pattern closely related to that in the human (Fig. 5D) and the macaque (Fig. 5E) . The expression of keratin 3 is weak throughout the eye epithelium during the first 4 days of embryonic development, (Figs 5F 5G) . In the adult zebrafish cornea, epithelial cells are strongly stained with the anti-keratin 3 antibody (Fig. 5H) . Very similar, if not identical, staining patterns are observed in adult human and macaque corneas (Figs. 5I 5J) . Corneal keratan sulfate proteoglycan is an excellent cornea-specific differentiation marker. Its expression is first detectable in zebrafish between 3 and 4 dpf (see Fig. 5K for 24 hpf; data not shown for 3 dpf). By 4 dpf, it is highly expressed in the corneal stroma (Fig. 5L) . No staining is detected in any other ocular tissues in zebrafish, macaque or human (Figs. 5L 5M 5N 5O) . In all immunostaining experiments presented, no signal is observed in zebrafish, macaque, or human corneas, after the omission of primary antibodies (data not shown). These results further illustrate the structural similarity between zebrafish and mammalian corneas.
|
|
| Discussion |
|---|
|
|
|---|
600 µm).23 Thus, even when we account for the differences in eye size, the zebrafish cornea is considerably thinner. The contribution of the stroma to corneal thickness is also different. In humans, the stroma accounts for
90% of corneal thickness (approximately 500 µm),23 much more than the 30% to 40% in the zebrafish. Another noticeable difference is the appearance of the corneal endothelium, which in the zebrafish seems to be less tightly organized. Scanning electron microscopy of the zebrafish endothelial surface reveals membrane folds. In addition, on TEM micrographs, junctional complexes are not obvious between corneal endothelial cells during the first 4 weeks of fish development. In contrast to that, we observed that the zebrafish corneal epithelial cells develop robust apical cell junctions already by 24 hpf. In adult fish, corneal epithelial cells are tightly connected to each other by numerous junctions. These observations suggest that the corneal epithelium, and not the endothelium, provides the major protective barrier in the fish cornea. Several key features of the larval and adult zebrafish cornea have been described previously.21 22 In this report, we extended previous studies by analyzing corneas from embryos and larvae at multiple stages of development. This analysis revealed the timing of key developmental events. We show that the morphogenesis of the zebrafish cornea starts by 20 hpf and possibly earlier. Similar to birds and primates,1 endothelial cells and keratocytes invade the cornea in two waves. The second wave, however, seems to be delayed in zebrafish, compared with those of birds and mammals. Although the migration and differentiation of the corneal endothelium in zebrafish is completed by 2 dpf, corneal keratinocytes are only observed between 14 and 28 dpf. A rapid increase in stromal thickness after the appearance of keratinocytes is noteworthy, as it suggests that this cell type is essential for stromal differentiation. In birds, corneal keratinocytes appear at stage 27 (55.5 days), immediately after the wave of endothelial migration that occurs at stage 22 (4 days).1 23 In humans and other primates, these two events are also separated by a very short interval, between stages 20 and 21.1 24 25 The second wave of migration in primates is replaced, however, by the differentiation of residual mesenchymal cells into keratocytes.25
Similar to that of mammals, the primary stroma of the zebrafish cornea acquires a regular organization only after the appearance of the endothelial layer, suggesting an endothelial contribution to the formation of the stroma. By contrast, the primary stroma in birds consists of approximately 30 orthogonally arranged strata of striated collagen fibrils before the cornea is separated from the lens by the endothelium.1 We also confirmed previous observations that lens formation in the zebrafish differs from that in mammals and birds.22 The morphogenesis of the zebrafish lens occurs through a delamination and rearrangement of cells in the lens placode, rather than by invagination. It will be interesting to determine at the level of genetic control what accounts for this difference.
The Zebrafish Model of Corneal Development and Its Genetic Underpinnings
Based on four lines of evidence, we propose that the zebrafish is an excellent model for studying corneal development, genetics, and disease. First, despite some notable differences, the zebrafish cornea is very similar in its anatomy and structure to the corneas of higher vertebrates, humans in particular. This conservation has been also suggested by other research groups.21 22 Second, several corneal proteins initially characterized in higher vertebrates are conserved in the zebrafish. Antibodies against these proteins display similar staining patterns in the zebrafish, human, and monkey corneas. Third, the zebrafish cornea develops very rapidly in embryogenesis, making it easily accessible to loss-of-function genetic analysis. Last, zebrafish mutants with corneal defects have been isolated (data to be published in a separate report), thus demonstrating that the zebrafish model is suitable for the study of corneal genetics.
The Zebrafish as a Model of Human Corneal Dystrophies
Corneal dystrophies are a heterogeneous group of frequently heritable disorders. The mechanisms that underlie their occurrence are poorly understood. The age of onset varies greatly for these abnormalities, from birth to the fourth decade of life.6 Scarring and other symptoms may be so severe that corneal transplantation is needed. Affected individuals experience discomfort in various forms, such as severe stabbing pain in the eye, foreign body sensation, and photophobia. The similarity of the zebrafish cornea to the human one makes the zebrafish an excellent model for studying human ocular diseases. In addition to the potential use of zebrafish mutant lines as models of human abnormalities, well-established genetics tools in zebrafish, such as transgenesis, permit the overexpression of genes carrying human mutations and the analysis of the resultant abnormalities. Similarly, the reverse genetic knockdown approach may be useful for the study of the developmental malformations of the cornea.
We have demonstrated in agreement with previous studies that the zebrafish cornea contains the Bowmans layer.21 22 This is an important observation, as the Bowmans layer plays a crucial role in corneal wound healing,26 and is affected by several human disorders, including Reis-Bücklers corneal dystrophy27 and keratoconus.28 Its presence in the mouse cornea is controversial,29 30 31 and thus the finding that Bowmans layer forms in zebrafish offers a strong argument for the relevance of the zebrafish model to human corneal dystrophies, particularly the ones that involve abnormal protein deposits in the Bowmans layer.
Our analysis indicates that the expression pattern of BIGH3, a gene responsible for several types of human corneal dystrophies, is conserved between zebrafish and higher vertebrates. In addition to expression pattern, the zebrafish gene shares a very high homology with its human orthologue, both at the DNA and protein levels.32 BIGH3 mutations commonly found in humans affect amino acid residues that are conserved between human and zebrafish. Therefore, the insertion of the BIGH3 gene containing human dominant mutations into the zebrafish genome via transgenesis is likely to generate corneal abnormalities related to human dystrophies. Such transgenic zebrafish lines can then be used to screen for chemicals that alleviate corneal defectsa particularly promising example of how the zebrafish model can be used to study and potentially cure a human corneal disease.
A Unique Opportunity to Study Corneal Differentiation and Maturation
Our analysis indicates that the zebrafish cornea displays a rapid early embryonic development, followed by a gradual differentiation and maturation of corneal layers. Critical processes in corneal differentiation and maturation include the formation of Bowmans layer and Descemets membrane, the invasion of keratocytes into the primary stroma, and the expansion of the secondary stroma. These developmental and differentiative events are not easily accessible to analysis in the mouse and the chicken, because they occur before birth in the mouse and before hatching in the chicken. In contrast to that, they are readily amenable to observation and molecular manipulation in zebrafish embryos or larvae, owing to their external development. The understanding of differentiative processes is important for the genetic and molecular studies of human corneal diseases, which can affect late-differentiating layers of the cornea.
Molecular Markers for Zebrafish Corneal Studies
We have identified several antibodies that are useful as markers of corneal development and differentiation in zebrafish. Anti-keratin 3 is an excellent marker of corneal epithelium. Corneal keratin sulfate proteoglycan is expressed specifically in the stroma, and thus an antibody directed to this antigen is a valuable tool for the analysis of stromal formation and differentiation. Both antibodies can also be used in corneal wound healing studies. Similarly, anti-collagen I antibody is likely to be a useful tool to study the interaction between the corneal epithelium and the stroma during development and wound healing. Further development of cornea-specific immunohistochemical markers, as well as the identification of zebrafish genes involved in corneal development via both forward and reverse genetic approaches will further strengthen the zebrafish as a model for the analysis of the basic molecular mechanisms of corneal development and their malfunction in corneal disease.
| Footnotes |
|---|
Submitted for publication December 17, 2005; revised June 15, 2006; accepted August 15, 2006.
Disclosure: X.C. Zhao, None; R.W. Yee, None; E. Norcom, None; H. Burgess, None; A.S. Avanesov, None; J.P. Barrish, None; J. Malicki, None
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked "advertisement" in accordance with 18 U.S.C.
1734 solely to indicate this fact.
Corresponding author: Xinping C. Zhao, Department of Ophthalmology and Visual Science, University of Texas Medical School at Houston, 6431 Fannin St., Houston, TX 77030; xinping.zhao{at}uth.tmc.edu.
| References |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |