SEARCH

SEARCH BY CITATION

Keywords:

  • eye field morphogenesis;
  • determination;
  • differentiation;
  • TH immunohistochemistry

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

This study describes a whole embryo and embryonic field analysis of retinoic acid's (RA) effects upon Xenopus laevis forebrain development and differentiation. By using in situ and immunohistochemical analysis of pax6, Xbf1, and tyrosine hydroxylase (TH), gene expression during eye field, telencephalon field, and retinal development was followed with and without RA treatment. These studies indicated that RA has strong effects upon embryonic eye and telencephalon field development with greater effects upon the ventral development of these organ fields. The specification and determination of separate eye primordia occurred at stage-16 when the prechordal plate reaches its most anterior aspect in Xenopus laevis. Differentiation of the dopaminergic cells within the retina was also affected in a distinct dorsoventral pattern by RA treatment, and cell type differentiation in the absence of distinct retinal laminae was also observed. It was concluded that early RA treatments affected organ field patterning by suppression of the upstream elements required for organ field development, and RA's effects upon cellular differentiation occur downstream to these organ determinants' expression within a distinct dorsoventral pattern. © 2001 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

The nervous system is a dorsalization of ectoderm when low levels of effective BMP-signaling result in the formation of the central nervous system (CNS). Subsequent to this “neuralizing” event in normal Xenopus embryos, the establishment of the characteristic anterior-posterior polarity of the neural tube is thought to be the result of a reinforcing two-step process. The nervous system is first specified to anterior fates and then a gradient of a transforming factor transforms neural tissue to posterior fates (Nieuwkoop, 1952). This model predicts that greater concentrations of this transforming factor should convert neural tissue to more posterior fates.

Vitamin A and related chemicals such as retinoic acid (RA) contribute to several aspects of the development of the olfactory area, the heart, respiratory tract, eye, and other embryonic structures that suggest its involvement in normal development and differentiation (LaMantia et al., 1993). A regulatory morphogenetic role of RA has also been suggested for the early developing nervous system, because exogenously applied RA produces a concentration- and stage-dependent truncation of anterior with enhancement of posterior neural structures in Xenopus embryos (Durston et al., 1989). At low concentrations, RA suppresses anterior development within the hindbrain, whereas at high concentrations, it causes more global anterior truncations of whole forebrain field areas. This suppression of anterior fields is reflected by RA's early suppression of anterior neural markers such as Otx (Papalopulu and Kintner, 1996).

The actual role of RA in anterior neural field development is still ambiguous and in need of study. Because most research has been based primarily on experiments that involve immersion of entire organisms in solutions of RA, the exact forebrain structures ablated and morphogenetic genes affected by RA-treatment are not known. Because several anterior structures are affected by whole-animal immersion (Durston et al., 1989), we were interested in RA's possible effects on the eye and the most anterior telencephalic structures. Therefore, we assessed the effects of RA treatment upon pax6-expression as well as RA's effects upon the telencephalon-specific Xbf1.

During neurulation, the presumptive eye tissue in the anterior neural plate separates into two eye primordia (Li et al., 1997). pax6-expression can be used as a marker for this eye field or primordium subdivision (Li et al., 1997). Before the eye fields bifurcate, pax6 mRNA is first expressed as a broad-banded, single eye-forebrain field at stage 12/13 during the beginning of neurulation (Li et al., 1997). The early expression of pax6 (Li et. al., 1997) and lineage dye experiments (Eagleson and Harris, 1990) indicated that, at stage-12, the eye field is a continuous band within the anterior neural plate and this expression band may subsequently split into two separate eye fields Eagleson et al., (1995). Li et al., (1997) by using neural plate explants and manually removing the prechordal plate, demonstrated that the prechordal plate mesoderm is important for the formation of two separate retinal or eye fields. Because these were explant experiments, these studies did not indicate specifically when the eye fields become separated, autonomous fields.

To determine precisely when the two eye fields become separated and the stage that the prechordal plate reaches its most anterior extent of the neural plate embryo, transplantations of median and lateral eye fields were done and ectopic expression of pax6 was followed. This assessed the progress of eye field separation and eventual organ field determination as two separate retinal (eye) fields.

Retinoic acid (RA) is also known to disrupt, ablate, or inhibit forebrain and eye (retinal) determination and development (Durston et al., 1989). In this study, this effect by RA was used to assess and follow the separated eye fields' determination and development. By looking at local and whole animal effects of RA during different early neural plate stages of development, one can precisely ascertain when the fields become completely separated and determined in vivo. It must be noted that in Xenopus laevis embryos, the anterior neural plate is completely underlain by mesoderm by stage 11/12 (Sharpe, 1992; Li et al., 1997). Thus, whole animal or local application of RA after neural plate stages should not affect underlying mesodermal movements of the neural plate (Sharpe, 1992). The prechordal plate merely elongates and extends along the midline (Keller et al., 1992) during later neurulation stages. By combining the results of these studies (pax6 and Xbf1 field determination and early effects of RA), we established when the two separate eye fields are autonomously determined.

We also began studies to investigate the effects of RA on cellular differentiation within eye fields looking at the development of the amacrine dopaminergic cells. These studies investigated the effects of RA treatment at different neurula stages upon dopaminergic cell differentiation in the dorsal and ventral aspects of the eye. We concluded that determination of these cells occurred during neural plate stages and before morphologic lamination of the retinae.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Effects of RA Treatment on the Telencephalon and Eye Fields of Neural Plate Stage Whole Embryos

Treatment of Stage-13 embryos for 2 hours (in 10−6 M RA) resulted in the complete absence of eyes, a cycloptic condition, or two partial eyes with retinal tissue restricted to the dorsal half of the eye (Fig. 1A). These RA-treated animals also lacked a pituitary, nasal pits, and telencephali. Treatment at stage-15 had a less drastic action on these forebrain structures. All tadpoles had eyes (although they were not fully formed), pituitaries, two much reduced nasal pits, and they lacked or had much reduced telencephali (Fig. 1C). Treatments at both stages resulted in greatly reduced anterior structures, but the later (stage 15) treatment resulted in less severe anterior truncations (compare Fig. 1C with 1B and with the control, 1D). RA primarily interfered with morphogenesis of the more anterior brain and sensory structures. As development proceeded more posterior, forebrain aspects (of even the diencephalon) became less affected (Fig. 2; Table 1). The ventral portion of the eye was more affected at earlier stages than later (compare Fig. 1C with 1B and with the control; 1D).

thumbnail image

Figure 1. Results of whole embryo immersion studies. A: Stage-13 (top organism) or stage-15 (middle organism) embryos were immersed into 10−6 M all-trans retinoic acid (RA) for 2 hr and then allowed to develop to stage 43/44 within amphibian medium. The bottom organism (A) was not immersed in RA. B–D: Ventral views of dissected brains from stage-13 RA-treated (B), stage-15 RA-treated (C), and untreated (D) tadpoles. Note in A that the animals immersed in RA at stage 13 lack a pigmented ventral portion of the eye. In the dissect brain (B), the eye has incompletely developed and resembles an undifferentiated optic vesicle (OV). In animals immersed in RA at stage 15 (A middle animal), the most ventral portion of the eye also lacks pigmentation (area between arrows; C). The animals that were immersed in RA at stage 15 also lacked much of the telencephalon (indicated by empty arrows), but had a pituitary (note; darkened animal in the middle of A). Animals that were not immersed in RA (A, bottom) had complete eyes that had pigmentation into the choroid fissure (arrow; D). Arrowtips point to the ventral portion of the eye affected by RA-treatment. Within C and D, arrows depict the telencephalon borders. Small arrow tips in C indicate the borders of the ventral area of the retina affected by RA treatment. p, pituitary (lacking in stage-13 RA-treated brains; B) OV, optic vesicle. Scale bars = 1 mm in A, 250 μm in D (applies to B–D)

Download figure to PowerPoint

thumbnail image

Figure 2. Anterior-to-posterior dose-dependent retinoic acid (RA) ablation of the embryonic frog brain. All treatments with RA were for 2 hr at room temperature (18-21°C) subsequent to the stage indicated. The most anterior portion of the brain (nasal areas/ olfactory bulb) were observed to be sensitive to 5 × 10−9 M RA; and other telencephalic areas were sensitive to 10−8 M RA. Higher concentrations 1 × 10−8 M to 5 × 10−7 M ablated diencephalic and even mesencephalic structures. SC, spinal cord; H, hindbrain; M, midbrain; D, diencephalon; T, telencephalon; pr, preoptic recess; inf, infundibulum.

Download figure to PowerPoint

Table 1. Effects of Different Retinoic Concentrations Upon Forebrain Structuresa
ConcentrationnDefect
Lack pituitaryLack eyesFused eyesIncomplete eyesLack noseSingle nose
  • a

    Stage 11/12 Xenopus embryos were immersed in RA diluted with 50% Holtfreter's solution for 2 hours. RA, retinoic acid.

Controls10000000
1 × 10−9 M RA15000500
5 × 10−9 M RA15200923
1 × 10−8 M RA15300837
5 × 10−8 M RA1540310510
1 × 10−7 M RA151028183
5 × 10−7 M RA151269093
1 × 10−6 M RA15144110150

Normal Expression Patterns of pax6 and Xbf1 Forebrain-Specific Genes

pax6 mRNA is initially expressed as a broad strip along the anterior border region of the neural plate at stage-12/13. Two broad strips of pax6 expression (not shown) are observed at stage 14, and by stage 15, presumptive lens ectoderm exhibits pax6 expression. Xbf1 mRNA is first observed as a narrow stripe along the most anterior region of the presumptive neural ridge area at stage 11 at the end of gastrulation (Fig. 3B). During neural plate stages, the most anterior extent of pax6-expression overlaps with the most posterior extent of Xbf1-expression (Fig. 3A,B; note small arrows as the anterior extent of Xbf1-expression). Additionally, there is Xbf1 mRNA detected within the anterior most portion of the retina in stage-28 embryos (Fig. 3C; arrow). This was the (neural ridge) area of Xbf1- and pax6-expression overlap in neural plate embryos.

thumbnail image

Figure 3. Whole embryo in situ hybridization for pax6 and Xbf1. A: A stage-15 embryo viewed frontally. The pax6-expression pattern is restricted to anterior lateral plate areas (including the lateral neural ridge) and a small bridge along the most anterior neural plate. pax6 expression is first detected at stage 12.5. The small arrow indicates the anterior extent of Xbf1 expression. B: A stage-13/14 embryo viewed ventrofrontally. The Xbf1-expression pattern (indicated by the small arrow) is within the anterior neural ridge tissue and a small portion of the anterior neural plate. The most ventral areas of Xbf1 expression overlap with the most anterior borders of pax6 expression, especially along the medial neural ridge. C: A stage-29/30 embryo. The Xbf1-expression pattern includes the most anterior prospective telencephalon, the branchial neural crest, and the dorsoanterior portion of the optic vesicle (arrow). Scale bars = 250 μm in A,C.

Download figure to PowerPoint

Effects of Whole Animal RA Treatment on pax6- and Xbf1 Expression

The distinct spatial region of Xbf1 expression within the presumptive telencephalic areas is disrupted by RA treatment at stages12 and 13 (Fig. 4A,B), but treatment of stage-14 embryos with RA does not significantly alter Xbf1 spatial expression at stage 28 (Table 2; Fig. 4C). pax6 expression within the presumptive optic vesicle regions was drastically altered by RA treatment at stages 12, 13, 14, and 15 (Table 2; Fig. 5A,B). RA treatment at stage 16 results in normal pax6 expression within the optic vesicles, but hindbrain pax6 expression is moved forward into the midbrain regions in RA-treated compared with untreated stage-28 embryos (Fig. 5C,D).

thumbnail image

Figure 4. The effects of whole embryo immersion in retinoic aid upon Xbf1 expression. A: At stage 12, embryos were placed in retinoic acid (RA) (10−6 M) for 2 hr and allowed to develop to stage 28. There is very little positive Xbf1 staining. This Xbf1 staining is rather diffuse, and optic vesicles did not evaginate. Much of the staining was probably nonspecific staining due to the open anterior brain cavity. B: At stage 13, embryos were placed in RA (10−6 M) for 2 hr and allowed to develop to stage 28. There was a small amount of Xbf1 expression (arrow) in the most anterior aspect of the embryo (presumptive telencephalon). These embryos had optic vesicles, but the anterior neuropore did not fully close. C: Embryos incubated in RA (10−6 M) at stage 14 for 2 hr exhibited normal Xbf1-expression patterns after it was allowed to develop to stage 28. These embryos also had normal optic vesicles. Scale bar = 250 μm in A (applies to A–C).

Download figure to PowerPoint

Table 2. Effects of Whole Animal Immersion in Retinoic Acid (RA) Upon Xbf1- and pax6-Expression Assessed at Stage 28a
Stage of immersion in 10−6 M RAXbf1 expressionpax6 expression
  • a

    pax6 expression differed from normal pax6 expression within all RA-treated stage-28 embryos (see Fig. 5C).

11/120/100/10
134/102/10
148/100/10
1510/102/10
1610/1010/10*
thumbnail image

Figure 5. The effects of whole-embryo immersion in retinoic acid (RA) upon pax6 expression. A: Embryos were placed in RA (10−6 M) at stage 12 for 2 hr and allowed to develop to stage 28. There is very little positive pax-6 staining and optic vesicles did not evaginate. The staining was probably nonspecific staining within the open anterior brain cavity. B: Embryos placed in RA (10−6 M) at stage 14 for 2 hr exhibited little pax 6 expression, despite that these stage-28 embryos developed normal optic vesicles. C: Embryos placed in RA (10−6 M) at stage 16 for 2 hr exhibited pax 6 expression at stage 28, but this expression pattern differed from stage-28 untreated embryos (D) in that hindbrain expression of pax 6 is anteriorized to within the midbrain (double arrows) in RA-treated embryos compared with normals (D). Scale bar = 250 μm in D (applies to A–D).

Download figure to PowerPoint

Local Effects of RA Treatment on the Morphogenesis of the Presumptive Eye Field

Experimental procedures are represented in Figure 6. Depending on the developmental stage of immersion in RA, the experimental presumptive eye areas (PEAs) either developed no recognizable eyes or had eyes that were partially or incompletely incorporated into the diencephalon of the brain. Retinoic acid (RA; 10−6 M) treatment of the presumptive eye field (PEF) at stage 11/12 resulted in the complete absence of an eye (Fig. 7A). The RA treatment of the PEA at stage 12/13 or later did not produce a significant effect on eye diameter (Table 2), even though eye morphogenesis was drastically affected (Fig. 7B). The major visible effect of RA treatment of the PEA during these later stages (12 through 15) was the appearance of a ventral area of the eye that lacked internal laminar organization and a pigmented retinal epithelium (arrowtips Fig. 8B,C). If present, the RA-treated eyes were only partially surrounded by the pigmented retinal epithelium usually at the dorsal aspect of the eye stem. Pigmentation was less pronounced in the ventral areas of those PEAs treated with RA (during stages 13–16).

thumbnail image

Figure 6. Experimental protocol for presumptive eye field studies. The prospective eye area (PEA) of an embryo previously treated 2 hr by immersion in 10−6 M all-trans retinoic acid at stage 11/12 (A) or stage 15 (B) was microdissected out, incubated 2 hr in 2 mg/ml of Hoechst 335 and then re-inserted into an embryo of the same stage but not immersed in RA. Embryos were allowed to heal and develop to swimming stages (43/44).

Download figure to PowerPoint

thumbnail image

Figure 7. Dorsal view of representative stage-43/44 tadpoles that had the prospective eye area (PEA) treated with RA at stage 14 (B) and stage 13 (C). A: An operated control, whereby the right PEA was removed, immersed in Hoechst solution, and reinserted. B: An animal that had its donor PEA treated with RA 10−6 M for 2 hr during stage 14. C: An organism that had its donor PEA treated with RA at stage 13 with 10−6 M for 2 hr and was reinserted. Note in B and C that the eye is attached to a nonpigmented diencephalon stalk with a portion of this “RA-treated” eye lacking pigmented retinal epithelium (arrows). These nonpigmented areas loop around to the ventral side and pigmentation of the eye is limited to the dorsal portion of the eye. Scale bar = 1 mm in C (applies to A–C).

Download figure to PowerPoint

thumbnail image

Figure 8. Effects of retinoic acid (RA) at stage 11/12 (A) and stage 14/15 (B) upon localized eye field morphogenesis. When embryonic prospective eye areas (PEAs) are treated at stage 11/12, this resulted in the complete lack of morphogenesis of a retina in many cases as seen by this dorsal view of the stage-43/44 tadpole. Arrows indicate Hoechst-positive tissue completely contained within the lateral diencephalon. Optic vesicle formation did not occur on the RA-treated side. When embryonic PEA is treated with RA at stage 14/15, the dorsal eye (retina) is normal, but the ventral area (arrow) exhibited the major visible effects of RA treatment by the appearance within the ventral area of the eye of a stub that lacked internal laminar organization and pigmented retinal epithelium. This is a ventral view of at stage-43/44 tadpole brain. pit, anterior pituitary. Scale bar = 500 μm in A; 100 μm in B.

Download figure to PowerPoint

Retinal morphogenesis was less suppressed in later stages (stages 14 to 16). A stem of “undifferentiated” tissue connected the ventral portion of the eye to the diencephalon in most cases (these ventral effects persisted with RA treatment up to stage 16).

Presumptive Eye Field Extirpation Studies to Assess Stage of Bilateral Eye Determination

Extirpation of the right presumptive eye area (PEA) during stages 12, 13, or 14 does not result in a significant loss in eye diameter when compared with the unextirpated side (data not shown). During stages 15 and 16, the presumptive eye area gradually loses its capacity to regulate for lost or extirpated eye field tissue (Fig. 9). By stage 16, over half (9 of 16) of the embryos failed to produce a visible eye when the PEA was extirpated at this stage (Fig. 9B).

thumbnail image

Figure 9. Effects of extirpation of the prospective eye area (PEA) during different neural plate stages. The right PEA was removed at stage 15 (A) and 16 (B). In the case of removal of the PEA in earlier stages, much of the eye area regenerated or was restored by stage 42. If the PEA is removed at stage 15 (A), only a small portion of the right eye is regenerated or restored. If the PEA is removed at stage 16, in 9 of 16 cases, no eye is regenerated by stage 42. Scale bar = 1 mm in B (applies to A,B).

Download figure to PowerPoint

Splitting and Determination of the Eye Fields in Whole Embryos

The experimental protocol for these studies is represented in Figure 10A. Midline and lateral neural plate tissue was dissected from one side and transplanted to the belly region on the same side of extirpation. This was done for stages 11/12, 13, and 16 (Nieuwkoop and Faber, 1967). The embryos were allowed to develop to stage 28 and were stained for pax 6 expression. Midline and lateral neural plate tissue from stage-11/12 embryos transplanted to the belly region was still capable of expressing pax 6 within both the belly graft and eye region in stage-28 embryos (Table 3; Fig. 10B). In fact, compared with other stages, a larger proportion of the graft expresses pax 6 when the operation is performed at this stage, and the eye region (origin of grafted tissue) was also capable of expressing pax 6 at stage 28. If the same experiment is performed at stage 15, midline anterior neural plate grafts to the belly do not express pax-6 (Table 3), but grafts from the lateral region during this stage do ectopically express pax-6. Interestingly, the lateral area source of these grafts did not express pax-6, even though it was the normal eye region (Fig. 10C; Table 3).

thumbnail image

Figure 10. pax6 expression in experimental embryos stained at stage 26. A: The experimental protocol. Midline areas (a,b,d,e,) were transplanted from neural plate staged embryos to the “belly” area on the same embryo. Also lateral areas (c,b,d,) were transplanted from staged neural plate embryos to the “belly” area of the same embryo (on the same side as the operated lateral area) and allowed to develop to stage 26, when they were stained for pax-6. B: A stage-26 embryo that had the midline area transplanted to its belly flank at stage a 11/12. Note that both the eye area and graft area stain positively for pax-6. C: A stage-26 embryo that had its lateral area transplanted to its belly flank at stage 15. Note that only the belly graft stains positively for pax-6, and the eye area lacks positive staining for pax-6. This is the same (lateral) side from which the stage-15 graft originated from. D: A stage-26 embryo that had the lateral area transplanted to its belly flank at stage 13. Note that both the eye area and the belly graft stain positively with pax6 (arrows).

Download figure to PowerPoint

Table 3. Effects of Retinoic Acid Upon Morphogenesis of the Eye in Xenopus laevisa
GroupnMean eye diameter (X ± SEM)
TreatedUntreated control
  • a

    Mean diameter is in micrometers. The presumptive eye region was dissected out of an embryo treated with 10−6 M RA for 2 hours and inserted into an untreated host embryo. Diameters compare treated eyes to their contralateral (untreated) eye. Due to the large variations in diameters, none of the groups were statistically different. In stage 11/12 and stage 12 groups some (n = 6) experimental groups completely lacked an eye, whereas others (n = 7) had very large eyes that lacked a pigmented retinal epithelium.

Stage 11/1214324 ± 89391 ± 50
Stage 1262347 + 123414 + 97
Stage 1313411 ± 96417 ± 96
Stage 1418393 ± 170440 ± 70
Stage 1524346 ± 134388 ± 60

Differentiation of the Dopaminergic Cells Within the Eye Field

To determine whether the action of RA was due to general inhibition of tissue development or the cessation of determination and subsequent differentiation, we looked at the development of specific cell types within the eye. This approach was accomplished by staining for the retinal dopaminergic amacrine cells. If PEAs were RA treated locally at different stages of development, this had a significant effect upon the number of TH-positive cells that differentiated within the eye or resultant optic stalk (Table 5). RA treatment significantly inhibited dopaminergic cell differentiation when PEAs were treated during stages 14, 15, and 16 (Table 4). This effect was more severe when the RA treatment was during earlier stages (Table 2). Interestingly, even stage-13 and stage-14 RA-treated PEAs later developed TH-positive cells within the ventral portion of their resultant stage-40/41 retinae (Fig. 11A). If one looked at the dorsal portion of the stage-14 RA-treated eye, TH-positive bipolar cells were observed within the amacrine layer of the dorsal retina (Figs. 11A, 12B, small arrows), whereas ventral TH-positive cells containing neural processes were scattered throughout the ventral tissue (Figs. 11B, left structure; small arrows; 12A) even though it was disorganized and lacked retinal lamination. If PEAs were treated with RA during stage-11/12, the resultant stage 40/41 had a few TH-positive cells contained in the grafted tissue area. The later neural stages that PEAs underwent RA-treatment resulted in greater number of TH-positive cells differentiating (Table 5).

Table 4. pax6 Staining at Stage 28 for Lateral and Midline Ectopic Areas of the Anterior Eye Field in Xenopus laevis
Stage of ectopic operation pax6 staining in graftpax6 staining in eye area (source of graft)
  1. aThe graft area staining was the ectopic belly region where the graft was transplanted. The “eye area” was the donor source of the (presumptive eye field) graft. + denotes intensity of staining in those grafts that stained; − denotes no staining.

Stage 11/12lateral+/− 1/9++
midline+ 5/9++
Stage 13lateral++ 6/8
midline+ 5/8++
Stage 15lateral+++ 8/8
midline+/− 2/8++
thumbnail image

Figure 11. Whole brain immunocytochemistry for tyrosine hydroxylase (TH) on brains in which the prospective eye area (PEA) was treated with RA. A,B: Examples of a brain which the PEA was treated at stage 14. Dorsal view (A) and a ventral view (B) of the same brain. A: Note that the right (RA-treated retina has a small mostly dorsal) portion that has undergone correct morphogenesis with retinal layering. This dorsal “laminated” portion has TH+ cells within the inner nuclear amacrine layer of cells (empty arrows). B: A ventral view of the same brain as A. Within the disorganized mass of cells of the ventral eye, there are located several TH+ cells, but this tissue lacks the organized lamination characteristic of a retina that has undergone proper morphogenesis. The TH+ cells have neural processes and lack the bipolar morphology observed within a properly laminated retina (arrows; control eye on right). C: A whole brain immunocytochemistry of a brain in which the PEA was treated for 2 hr with RA during stage 13. The “eye-structure” that resulted was an extended but large eye stalk (or optic vesicle; ov). This structure lacked any internal lamination and had not undergone complete morphogenesis subsequent to optic vesicle formation. Despite this lack of internal morphogenesis, a few TH+ cells were observed within the stalk of RA-treated structure. Most of the TH+-positive cells were localized near the stalk connecting the stalk to the diencephalon. TP, nucleus of the tuberculum posterior. Scale bar = 250 μm in C (applies to A–C).

Download figure to PowerPoint

thumbnail image

Figure 12. Whole brain immunocytochemistry for tyrosine hydroxylase (TH) on brains in which the prospective eye area (PEA) was treated with RA. A,B: Magnified views of Figure 11B. C: Magnified view of Figure 11C. A: Ventral view of PEA region that was retinoic acid (RA) treated at stage 14 and allowed to develop to stage 42. The ventral region contains TH-positive cells (top arrow) scattered throughout the nonlaminated portion of this structure. TH+-cells at the edge to the laminar portion exhibit neural processes. B: Ventral view of a control, untreated eye. Arrows indicate bipolar cells on the amacrine layer of the retina. C: Ventral region of a PEA region RA treated at stage 13.The tuberculum posterior (TP) area stain positively for TH and individual cells along the optic vesicle (OV) also stain. Many (top arrow) have neuronal processes. Scale bar = 125 μm.

Download figure to PowerPoint

Table 5. Number of Tyrosine Hydroxylase–Positive Cells (mean ± SEM) Within the Eyes of Stage-41 Xenopus Tadpolesa
Stage or RA-treatmentnControl eyeRA-treated eye
  • a

    RA, retinoic acid; n = number of brains assessed.

  • *

    Significantly different: *P < 0.01; **P < 0.01; t-test.

Stage 16630.8 + 1.1617.4 + 1.86*
Stage 15833.7 + 5.6611.8 + 3.65*
Stage 13829.8 + 4.104.8 + 2.37**

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

RA-treatment at neural plate stages had its most obvious effects upon the anterior brain regions. This included truncation or complete ablation of the telencephalon, eyes, nasal areas, and pituitary (Table 1). These were all areas derived completely (or in part) from the anterior neural ridge of the stage-15 neural plate embryo (Eagleson and Harris, 1990). As neurulation proceeded, later RA-treatments resulted in less extensive truncation of anterior structures (Figs. 1, 2). The telencephalon and ventral eye were the obvious structures most affected by RA-treatment during or after stage 15 (Figs. 1, 2). Preliminary studies also indicated that these are the areas with the greatest sensitivity to lower concentrations of RA (10−8 to 10−9 M; Fig. 2; Table 1).

The early neural expression genes of Otx1 and Otx2 are known to control certain aspects of early forebrain morphogenesis (Acampora and Simeone, 1999). Otx2 is necessary for specification and maintenance of anterior neural plate and late gastrulation events (Acampora and Simeone, 1999). In Xenopus embryos, Otx2 is responsive to RA and important for inductive interactions between rostral neuroectoderm and axial mesendoderm (Ang et al., 1994). Thus, the early exposure of Xenopus embryos to RA during stages 11/12, might disrupt Otx2 maintenance of the early presumptive forebrain tissue interactions necessary of early regionalization events, whereas RA exposure during later stages could suppress or affect genes downstream to these events (e.g., Xbf1 and pax6). These later genes are involved in specific forebrain organ regionalization and later organ-specific cellular specification events (Li et al., 1997; Bourguignon et al., 1998).

Studies investigating the effects of RA on expression of the telencephalon marker (Xbf1) and the eye marker (pax6) indicated that RA affects these forebrain tissues by suppressing initial gene expression cascades involved in organ field development and regional fixation or specification (Figs. 2–4). If embryos were treated with RA at stage 11/12, this completely suppressed later (stage 28) expression of Xbf1 and pax6, possibly by acting on the earlier Otx2-6 responsive elements. RA treatment at this stage (11/12) also suppressed optic vesicle formation (Fig. 4). RA treatment at stage 13 suppressed pax6 expression, but Xbf1 expression was not suppressed (Fig. 5). Interestingly, RA treatment during stage 13 did not suppress optic vesicle formation, despite its suppression of pax6 expression. Xbf1 is important for telencephalon vesicle formation (Xuan et al., 1995) and is expressed within overlapping borders of the telencephalic and eye fields (Figs. 3–5). BF-1 has also been demonstrated to be important for dorsal eye development in mice (Huh et al., 1999). Therefore, we concluded that Xbf1 may be important in optic vesicle formation as well as telencephalon vesicle formation and is a downstream regionalization event to Otx2 action. Thus, this early effect of RA on eye organogenesis could result in the lack of an eye when treated at stage 11/12 (Fig. 7A) because of RA's inhibition of optic vesicle formation. RA primarily affected ventral eye organogenesis during later stages. These later effects did not significantly alter eye size (Table 2; Fig. 8) but primarily prevented retinal lamination thus disrupting downstream morphogenetic events.

Asymmetry of the eye along the dorsal-ventral (D-V) axis is most evident morphologically by the presence of the ventral choroid fissure of the eye. D-V axial information in the eye is established very early in development as demonstrated by the studies of Papalopulu and Kintner (1996), whereby the Xbr1 gene that marks the dorsal ciliary margin is expressed as early as neural plate stages (within the neural ridge area). Several other markers have been identified that distinguish between dorsal and ventral retina (Constantine-Paton et al., 1986), but the role these molecules play in D-V patterning remains to be established.

In zebrafish and Xenopus (March-Armstrong et al., 1994), RA levels are higher in the ventral half of the embryonic eye. The present study demonstrated that throughout later neural plate stages, the ventral half was more sensitive to exogenous RA. In zebrafish and Xenopus, the retinoic acid synthetic enzymes are also asymmetrically distributed along the D-V axis (Wagner et al., 2000). RA applied to the presumptive eye area (PEA) at stages when the optic vesicle begins extending out (Eagleson et al., 1995) has its greatest effect on D-V morphogenesis by its disruption of the ventral aspect of the eye (Figs. 6, 7). The morphogenesis of the dorsal aspect of the retina (observed as laminated structure) is less affected by exposures to RA during later stages. After stage 16, the retina becomes less susceptible to RA-induced morphologic alterations. In addition, extirpation of the eye area (PEA) at this stage resulted in a loss of the ability to regenerate lost eye structures (Fig. 9). This desensitization to RA is probably due to increased endogenous RA and its previous actions on morphogenesis, especially in the ventral aspect of the eye (Wagner et al., 2000). The desensitization to RA and the loss of the ability to replace lost presumptive eye tissue strongly indicated that these morphologic field events are “determined” or specified and, therefore, less affected by exogenous factors after stage 16.

The period of RA sensitivity precedes evidence of gross morphologic or histologic differentiation of the eye by at least 12 hours or 10 stages (Nieuwkoop and Faber, 1967), suggesting that the mechanisms controlling the organization of eye structures are already in progress during these late neural plate stages. This determination of morphogenetic processes was also well in advance of differentiation of retinal cell types.

To test the eye-forming field potential of regions of the anterior neural plate during different stages of neurulation, we transplanted midline and lateral areas to the belly of the embryo and allowed them to develop to stage 28 when they were tested for pax6 expression. These studies demonstrated that both midline and lateral areas of early embryos (stages 11/12 & 13) were capable of ectopically expressing pax6. After stage 15, the midline area was less capable of ectopic expression of pax6 (Table 3). Lateral regions from stage-15 embryos were capable of ectopically expressing pax6, but the source of the explant was not capable of compensating for the lost tissue, and the normal eye area did not express pax6 at stage 28. These studies demonstrated that, at stage 12, the eye field is a continuous band within the most anterior neural plate. The eye field subsequently splits and regionalizes into two specific retinal regions by stage 15/16. These two retinal fields become separated by the anterior extension of the prechordal plate (Li et al., 1997) as it reaches its most anterior extent by stage 16. That these retinal fields are fully specified is further indicated by the extirpation studies (Fig. 9), whereby extirpation of the lateral area at stage 16 resulted in the complete loss of the eye in animals allowed to develop to stage 42. Thus, RA experiments, in situ hybridization for pax6 of eye-field explant studies and extirpation experiments all confirm that separated, specified retinal eye fields are determined at stage 16.

Cellular differentiation within the retina was also affected by RA treatment at different stages of development. Only a few TH-positive cells develop within PEA tissue that was treated at stage 11/12 (initial formation of the neural plate). During later stages, TH-positive cells can be detected within the retina, but the number of TH-cells within RA-treated PEAs becomes greater as neurulation proceeds. Pituello et al., (1989) also found that neuronal precursors became committed at early neural plate stages of neurogenesis. Greater commitment to TH-like fates during later neural plate stages indicated that the presumptive-TH cells become progressively resistant to RA's effects during later stages of development. It also indicated that this cell type became “determined” during neurulation, even though its differentiation did not occur until much later (Huang and Moody, 1992).

Recent studies by Ye et al. (1998) indicated that signals from the anterior neural ridge (ANR) and the prechordal plate may influence the distribution of the dopaminergic cells within the rostral forebrain. Because these areas express these signals primarily during neurulation, RA may be disrupting these signal interactions during early neurulation and suppressing subsequent differentiation of dopaminergic cells within the eye. It is of interest that TH-positive cells can be observed within the ventral retina even though the morphogenesis (lamination) of this area has been disrupted by RA (Figs. 11, 12).

This study demonstrated that RA disrupts the morphogenetic transformation of a part of the diencephalon into the eye with stronger influences on the area that ultimately becomes the ventral retina. The specification of separate eye primordia was ascertained to occur at stage 16 when the prechordal plate reaches its most anterior aspect. Differentiation of the retina is also affected by RA, but cell types found within distinct layers of the retina may differentiate to the exclusion of proper morphogenesis of these layers.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Embryos

Eggs and embryos were obtained by injection of human chorionic hormone (Sigma). Embryos were kept at 16°C in Steinberg's (Rugh, 1961). Embryos were manually dejellied by using Dumont forceps (No. 5).

In Situ Hybridization

Whole-mount in situ hybridization was performed according to Harland (1991), with minor modifications. Antisense RNA probes were labeled with digoxigenin-UTP or fluorescein-UTP by using the Boehringer (Roche) Genius RNA labeling kit. RNA localization was visualized with BCIP/NBT. The full-length Xbf1 probe was generously provided by Dr. Nancy Papalopulu, and the pax6 probe was generously provided by Dr. Rao.

Embryo Manipulations

Embryos were fertilized, dejellied, and raised according to Eagleson and Harris (1990) and staged according to Nieuwkoop and Faber (1967). The neural stages used in these studies included 11/12, 13, 14, 15, and 16. Once embryos reached a specific stage, they were immersed in all-trans retinoic acid (Sigma) at concentrations of 10−5 or 10−6 M for 2 hr. Whole embryo studies were performed by first immersing in RA and then allowing the embryos to develop to stage 28 or 40 in Steinberg's solution. To determine the effects of RA on telencephalic (Xbf1) or eye field (pax6) specific genes, the stage-28 animals were subjected to in situ hybridization for Xbf1 or pax6.

For studies on localized effects of RA, embryos were transferred to Steinberg's solution and the presumptive eye area contained within neural ridge tissue (Eagleson and Harris, 1990) was microdissected out by using Minuitin pins or electrosharpened tungsten needles. This presumptive eye region was microdissected out during stages 13, 14, and 15 and included only the neural ridge or presumptive ridge area. Because after the 2-hr treatment the embryo had not reached a stage with visible neural ridges, stage-11/12 embryos had prospective neural ridge and adjacent anterior plate tissue microdissected out. The microdissected tissue was placed in Hoechst dye (2 mg/ml in Steinberg's solution) and re-inserted into an untreated host embryo of the same stage as the donor embryo. Care was taken to maintain the original tissue orientation in the new host. Operated and control (Hoechst-labeled but not RA-treated) tadpoles were allowed to develop normally until they reached stage 44/45. Stage 44/45 tadpoles were fixed in 4% paraformaldehyde and stored until used.

To investigate the stage that presumptive eye regions were no longer capable of replacing lost tissue areas, presumptive eye extirpation operations were performed. During these studies, embryos at stage 11/12, 12, 13, 14, 15, and 16 had the presumptive eye region microdissected out (as previously described) and the tissue was not replaced.

Eye Measurements

Because localized RA treatment at stage11/12 often resulted in the complete or partial loss of an eye structure, it was thought that inhibition or ablation of the eye field might be a mechanism of RA's teratogenic effects. To test this possibility, eye size was measured to ascertain whether decrease in size was a major feature of RA treatment upon eye morphogenesis. Tadpoles and staged micrometers were photographed by using an Olympus SZH-10 dissecting microscope. By using this staged micrometer, the diameter of the eye could be measured by enlarging the negative onto a piece of paper and drawing the magnified eye with the stage micrometer. Actual areas could be calculated by using a planimeter. The operated eye was then compared with its contralateral untreated eye.

The Splitting of the Eye Fields in Whole Embryos

To follow the timing of the full extension of the prechordal plate and the complete splitting of the eye fields, embryos were exposed to RA during different neural plate stages. The lateral or median anterior neural plate (presumptive eye region) was ectopically implanted to the belly of the same operated individual. Embryos were allowed to develop to stage 28, and they were probed by means of in situ hybridization for pax6 expression.

Differentiation of the Eye Dopaminergic Cells

To test for differentiation of anterior markers, tyrosine hydroxylase (TH) immunocytochemistry was used. TH-positive neurons are found extensively within the forebrain (Tuinhof et al., 1994) and the eye (Huang and Moody, 1992). Fixed whole-brains with adhering eyes of RA-treated and untreated tadpoles were dissected out. To identify the grafted tissue(s), the dissected brains were observed and photographed under a Zeiss Standard-14 microscope equipped with epifluorescence. The brains were transferred to phosphate-buffered saline (PBS; pH 7.4), then to PBS with 0.2% bovine serum albumin and 0.1% Triton-X and whole brain immunocytochemistry for TH was done as previously described (Eagleson et al., 1998).

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

The authors thank Mary “Lucy” Wentworth for her excellent technical help throughout this study. We would also thank Professor Antone G. Jacobson for commenting upon early drafts of this study. G.W.E. was supported by an AREA NIH Grant.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES