Truncated Forms of Pax-6 Disrupt Lens Morphology in ... - CiteSeerX

Truncated Forms of Pax-6 Disrupt Lens Morphology in Transgenic Mice Melinda K. Duncan,1 Ales Cvekl,2 Xuan Li,3,4 and Joram Piatigorsky3 PURPOSE. Extensive literature shows that Pax-6 is critical for lens development and that Pax-6 mutations can result in aniridia in humans. In addition, it has been reported that truncated Pax-6 molecules can act as dominant–negative repressors of wild-type Pax-6 activity in cultured cells. This study was designed to determine whether Pax-6 molecules without either the activation domain (AD) or the homeodomain (HD) and the AD can function as dominant–negative repressors in vivo and alter the phenotype of the lens. METHODS. Transgenic mice were created harboring the ␣A-crystallin promoter linked to a cDNA encoding either a truncated Pax-6 without the C terminus (paired domain [PD] ⫹ homeodomain) or Pax-6 consisting of only the PD. The phenotype of the resultant animals was investigated by light and electron microscopy as well as atomic absorption spectroscopy. RESULTS. Two lines of PD ⫹ HD mice and three lines of PD mice were generated, all of which exhibit posterior nuclear and/or cortical cataracts of variable severity. The lenses from mice transgenic for either Pax-6 truncation are smaller and more hydrated than normal. Morphologically, the mice expressing the PD ⫹ HD of Pax-6 have swollen lens fibers with attenuated ball-and-socket junctions. In contrast, the lenses from mice overexpressing the PD of Pax-6 have posterior nuclear cataracts composed of cell debris, whereas the remaining fiber cells appear generally normal. CONCLUSIONS. The presence of truncated Pax-6 protein in the lens is sufficient to induce cataract in a wild-type genetic background. The simplest explanation for this phenomenon is a dominant– negative effect; however, a number of other possible mechanisms are presented. (Invest Ophthalmol Vis Sci. 2000;41:464 – 473)

A

niridia is a semidominant disease characterized by the absence of iris development with associated corneal opacity.1 Patients with this disorder also often have glaucoma, defects in the anterior chamber angle, and cataract.2 Aniridia is usually caused by mutations in the gene encoding the transcription factor, Pax-6.3 Humans and mice homozygous for Pax-6 mutations exhibit multiple developmental defects4,5 because this gene is required for the induction of the lens from the embryonic ectoderm, the development of the nasal placode,6 and the patterning of the forebrain7 as well as the development and maintenance of pancreatic acinar cells8 and the corneal epithelium.9

From the 1The Department of Biological Sciences, The University of Delaware, Newark; 2The Departments of Ophthalmology and Molecular Genetics, Albert Einstein College of Medicine, Bronx, New York; and the 3Laboratory of Molecular and Developmental Biology, National Eye Institute, Bethesda, Maryland. 4 Present address: OriGene Technologies Inc., 6 Taft Ct. Suite 300, Rockville, MD 20850. Supported by a grant from the Knights Templar Eye Foundation (MKD), Grant RO1 EY12221-01 from The National Eye Institute (MKD), a Research to Prevent Blindness Career Development Award (AC), intramural research support (1Z01EY00126 –17) from The National Eye Institute (JP); and National Institutes of Health IDEA Grant RR11820 to the University of Delaware. Submitted for publication May 20, 1999; revised August 16, 1999; accepted September 23, 1999. Commercial relationships policy: N. Corresponding author: Melinda K. Duncan, Department of Biological Sciences, The University of Delaware, 331 Wolf Hall, Newark, DE 19716. [email protected]

464

The Pax-6 protein consists of at least three functional domains. The paired domain (PD), named for its similarity to the Drosophila protein Paired, consists of two subdomains (PAI and RED), each of which contribute to the DNA binding specificity of Pax-6.10,11 The homeodomain (HD) functions as a dimerization as well as a DNA-binding domain that can act independently and in cooperation with the PD to increase the spectrum of possible DNA-binding sites.10 The C terminus of Pax-6 is rich in proline, serine, and threonine amino acids and is thus denoted the PST domain. The PST domain is critical for the full range of Pax-6 function and appears to be a transcriptional activation domain (AD).5 More than 70 different mutations affecting the eye have been identified in the Pax-6 gene of vertebrates, ranging from the complete deletion of one allele12 to missense mutations.13 Different eye tissues have different sensitivity to alterations both in the amount of functional Pax-6 protein present14 and in Pax-6 protein structure.11,15 The most commonly identified mutations result in the premature termination of mRNA translation and in the production of truncated Pax-6 proteins.3 In patients with classic aniridia, the severity of the iris and corneal pathology does not correlate with the type of Pax-6 mutation, suggesting that the disease is caused by a haploinsufficiency of Pax-6 function.1 However, biochemical characterization of mutant Pax-6 proteins indicates that some can maintain partial function.5,16 –18 Early in development, Pax-6 is critical for the induction of the lens from the head ectoderm.6 Later in lens development, Pax-6 appears to transactivate the expression of a number of crystallin genes, including ␣A-,19 ␣B-,20 ␦-,21 and ␨-crystallin,22 Investigative Ophthalmology & Visual Science, February 2000, Vol. 41, No. 2 Copyright © Association for Research in Vision and Ophthalmology

IOVS, February 2000, Vol. 41, No. 2

Truncated Forms of Pax-6 Can Disrupt Lens Morphology

and to repress the expression of the lens fiber cell–specific ␤B1-crystallin gene.17 The importance of Pax-6 in the lens after induction can be inferred from the observation that aniridia patients commonly develop cataracts.2 Although it has been suggested that these cataracts arise secondary to the other eye pathologies of aniridia patients,3 a family has been reported with relatively normal eye anatomy except for the presence of bilateral congenital cataract. In this family, a mutation results in the deletion of the last 69 amino acids of the Pax-6 AD.5 This is consistent with the observation that aniridia patients with mutations in the AD of Pax-6 are more likely to suffer from congenital cataracts than patients with other types of Pax-6 alterations.23 If cataracts are common in aniridia patients that express truncated Pax-6 proteins, it is possible that the mutant protein interferes with the function of protein expressed from the normal allele. This hypothesis is supported by the fact that truncated Pax-6 proteins bind DNA with higher affinity than wild-type Pax-6 and can act as dominant–negative repressors of Pax-6 function in tissue culture.18 Thus, we tested whether truncated forms of Pax-6 could interfere with lens development and/or function in vivo, by using transgenic mouse technology.

MATERIALS

AND

METHODS

Transfection Analysis The plasmids pG5ECAT (containing the E1b minimal promoter driving the CAT reporter gene) and a derivative plasmid containing 6 copies of the Pax-6 PD binding site cloned upstream of the E1b minimal promoter in the E1b vector have been described elsewhere11 (a gift from Richard Maas, Harvard Medical School, Boston, MA). The parental expression plasmid pKW10, pKW10 expressing Pax-6 (pPax-6), pKW10 expressing the PD and HD of Pax-6 (the N-terminal 286 amino acids; pP6⌬286), and pKW10 expressing the PD of Pax-6 (the Nterminal 140 amino acids; pP6⌬140) have been described previously.17 pCMV␤Gal was purchased from Clontech (Palo Alto, CA). N/N1003a cells, an established rabbit lens epithelial cell line, (a gift from John Reddan, Oakland University, Rochester, MI) were maintained as described.24 ␣TN4-1 cells, a T-antigen– transformed mouse lens epithelial cell line, were maintained as described.25 Chinese hamster ovary (CHO) cells, an established fibroblast cell line, (a gift from Ulhas Naik; The University of Delaware) were maintained as described.26 3T3-Tag, a T-antigen–transformed version of the mouse fibroblast cell line 3T3, (a gift of Daniel Simmons; The University of Delaware) were maintained as described.27 For all transfections, all cells were plated at a density of 7.5 ⫻ 105 per 60-mm dish. The next day, the cells were fed fresh media, and transfections were performed with 10 ␮g promoter/CAT plasmid, 1 ␮g pCMV␤Gal, and various amounts of expression vector using cationic liposomes (Lipofectamine; Life Technologies, Gaithersburg, MD) as described.28

Production of Transgenic Mice The plasmid pACP2 (a gift of J. Fielding Hejtmancik, National Eye Institute) containing the mouse ␣A-crystallin promoter (⫺342/⫹49),29 the simian virus (SV)40 small T-antigen intron and polyadenylation signal were modified by the addition of a

465

polylinker downstream of the ␣A-crystallin promoter containing ClaI, BclI, XhoI, ApaI, HindIII, SpeI, and BglII sites and transformed into DM1 competent cells (Life Technologies, Gaithersburg, MD) to create the plasmid pACP3. Pax-6 cDNAs (full length), PD ⫹ HD (amino acid [aa] 1–286) and PD (aa 1–140) were removed from pPax-6, pP6⌬286, or pP6⌬140 by digestion with BamHI and HindIII and were then ligated into the BclI–HindIII site of pACP3. The fragment containing the ␣A-crystallin promoter, the Pax-6 coding sequences, and the SV40 small T-antigen intron and polyadenylation sequence was liberated from the resultant plasmids and purified using glass milk.30 The resultant fragment was used by the National Eye Institute transgenic mouse facility to generate transgenic mice (strain FVB/N) by pronuclear injection of fertilized eggs, as described.30 Transgenic animals were identified by PCR analysis of DNA obtained by tail biopsy using primers generated from the SV40 small T-antigen intron as described.31 All experiments using animals were approved by the National Eye Institute and University of Delaware institutional review boards and conformed with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research.

Gross Morphology Mice were killed by cervical dislocation and the eyes enucleated and fixed in 4% neutral buffered formalin. After three hours, the lens was removed and photographed with a dissecting microscope (Stemi SV11 apo; Carl Zeiss, Thornwood, NY) fitted with a Pixera digital camera (Pixera, Los Gatos, CA) and ring light illumination.

Histologic Analysis Mice were killed by cervical dislocation or decapitation, and the eyes enucleated and transferred to 4% neutral buffered formalin. After 18 hours of fixation, the eyes were transferred to 70% ethanol and stored until paraffin embedding. Six-micrometer-thick sections were prepared and stained with hematoxylin and eosin by standard methods.

Immunocytochemistry The expression and localization of truncated Pax-6 proteins in the lens was determined by indirect immunofluorescence using a pan-specific PD primary antibody (1:1000 dilution, Bios, Prague, The Czech Republic) and rhodamine red X–labeled anti-rabbit IgG (Jackson ImmunoResearch, West Grove, PA) on paraffin-embedded lens tissue from 12-week-old adult mice. The resultant fluorescence was detected on a confocal microscope (510 LSM; Zeiss) configured with an argon-krypton laser (488 nm and 568 nm excitation lines).

Lens Protein Analysis Mice were killed by cervical dislocation, the eyes enucleated and lenses dissected. Both lenses from the same animal were homogenized in 200 ␮l of a buffer containing 20 mM sodium phosphate and 1 mM EGTA (pH 7.0). The soluble and insoluble fractions were separated by centrifugation and the concentration of the soluble proteins determined by the BioRad protein assay (Bio-Rad, Hercules, CA). The crystallin profile of lenses was determined by electrophoresing 30 ␮g soluble protein on a 15% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) gel followed by staining with Coomassie blue. Western blot analyses were performed by electrophoresis

466

Duncan et al.

of 100 ␮g soluble protein or 10 ␮g ␣TN4-1 cell extract on a 12% SDS-PAGE gel, transfer of the separated protein to nitrocellulose, and immunodetection of Pax-6 PD containing proteins using the anti-PD antibody at 1:5000 dilution followed by detection with enhanced chemiluminescence (Amersham– Pharmacia, Piscataway, NJ).

Scanning Electron Microscopy Mice were killed by cervical dislocation, the eyes enucleated and transferred to 0.08 M sodium cacodylate, 1.25% glutaraldehyde, 1% paraformaldehyde (pH 7.3). After several hours of fixation, lenses were excised from the eyes and placed in fresh fixative for 48 hours. After fixation, the lens capsule and outermost layer of fiber cells were dissected from the lens. The peeled lens was transferred to 70% ethanol and incubated overnight. The lens was dehydrated by two 5-hour incubations in 100% ethanol and dried in 1:2 hexamethyldisilazane (HMDS; Sigma, St. Louis MO)/ethanol for 1 hour, 2:1 HMDS-ethanol for 1 hour, and two changes of 100% HMDS for 30 minutes each. The lenses were subsequently transferred to filter paper and placed in a vacuum dessicator until analysis. The lenses were mounted on stubs with double-sided tape and sputter coated for 2 minutes with gold-palladium. Specimens were visualized with a scanning electron microscope (JEOL, Peabody, MA).

Lens Hydration and Calcium Content Mice were killed by cervical dislocation, eyes enucleated and lenses dissected. After vitreous and aqueous humor were blotted, the lenses were weighed and left at 100°C for 24 hours. The percentage of hydration was calculated after determination of dry weight. Each dry lens was digested by adding 50 ␮l concentrated HCl and incubating for 3 days at room temperature. The sample was then diluted to 500 ␮l with distilled, deionized water and the calcium concentration determined by atomic absorption spectroscopy (model 800; Perkin Elmer, Norwalk, CT).

Small Eye Mice Heterozygous mice harboring the Sey⬍Dey⬎ mutation were obtained from the Jackson Laboratory (Bar Harbor, ME) and maintained in the laboratory by backcrossing males to females of the strain FVB/N. Mice used in the present study have been crossed to FVB/N for 4 to 5 generations. Eyes from mice harboring the Sey⬍Neu⬎ mutation were a gift from Yasuhide Furuta and Brigid Hogan, Vanderbilt University (Nashville, TN).

RESULTS Truncated Pax-6 Proteins as Dominant–Negative Repressors in Transfected Lens Cell Lines Recently, truncated Pax-6 proteins were shown to be capable of blocking Pax-6 –mediated transcriptional activation in transfected mouse fibroblasts.18 Because this finding has implications for the ocular phenotypes in aniridia, we tested the ability of truncated Pax-6 molecules to repress Pax-6 – mediated transcriptional activation in the lens-derived cell lines N/N1003a (Fig. 1A) and ␣TN4-1 (Fig. 1B). In both these cell lines, which were previously shown to express Pax-6 endogenously,19,32 the presence of six Pax-6 PD-binding sites upstream of the E1b promoter (6XPax-6 con) resulted

IOVS, February 2000, Vol. 41, No. 2 in a large increase in CAT protein expression. Increasing the levels of Pax-6 in these cell lines by cotransfection of a Pax-6 expression plasmid resulted in a repression of reporter gene expression consistent with previous reports.21 Cotransfection of an expression vector producing truncated Pax-6 proteins consisting of either the PD and HD (aa 1–286) or PD alone (aa 1–140) was much more efficient than fulllength Pax-6 in repressing the expression of the reporter vector containing six Pax-6 – binding sites. To ensure that the observed repression was caused by direct interference with wild-type Pax-6 function, these transfections were repeated in two fibroblast cell lines that do not express Pax-6. In CHO (Fig. 1C) and SV40 T-antigen– transformed 3T3 (Fig. 1D) cells, the presence of six Pax-6 consensus PD-binding sites did not affect the activity of the reporter gene relative to the control plasmid containing the E1b promoter alone. Cotransfection of 6XPax-6 con with various amounts of Pax-6 expression vector in both CHO and T-antigen–transformed 3T3 cells activated reporter gene expression five- to sevenfold, whereas expression vectors producing either of the truncated forms of Pax-6 used in this study had no effect. Cotransfection with 500 ng Pax-6 expression vector with PD ⫹ HD expression vector led to a 60% to 80% reduction in reporter gene activity. In contrast, the transcriptional activation of the reporter gene by Pax-6 was not significantly reduced by cotransfection with the PD (aa 1–140) expression vector in either cell line. This differs from the repressive effect of PD in the lens cell lines described earlier.

Cataracts in Mice Transgenic for ␣A-Crystallin– Truncated Pax-6 Transgenes Because both truncated forms of Pax-6 used in the cotransfection experiments were able to repress transcriptional activation mediated by Pax-6 PD-binding sites in lens cell lines, and because patients with cataract associated with aniridia have been shown to harbor Pax-6 genes that could give rise to similar proteins,3,33 we tested the ability of truncated Pax-6 alleles (PD ⫹ HD, aa 1–286; PD, aa 1–140) to disrupt lens function in a wild-type genetic background using transgenic mice. For these experiments, the ␣A-crystallin promoter was used, because it is well established that it directs moderate to low levels of transgene expression specifically to lens fiber cells.29,34 Two independent lines of ␣A-PD ⫹ HD and three lines of ␣A-PD mice were generated. One line of ␣A-PD ⫹ HD mice had severe cataracts (Fig. 2B), whereas the other line had a milder phenotype (data not shown). Two lines of ␣A-PD mice had posterior nuclear cataracts (Figs. 2C, 2D). The line of ␣A-PD mice with clear lenses at weaning had opacities develop by 16 weeks of age (data not shown). The correlation between the genotypes and phenotypes described was 100% in all cases. Nontransgenic littermates were never observed to have any lens abnormalities.

Light Microscopy The wild-type mice used as control subjects in this experiment maintain clear lenses until at least 12 months of age, with no alterations notable histologically (Fig. 3A). Twelve-week-old mice expressing the PD ⫹ HD construct had small lenses with swollen cortical fibers and an unevenly stained lens nucleus (Fig. 3B). Twelve-week-old mice expressing the PD construct



467

FIGURE 1. The ability of truncated Pax-6 proteins to repress Pax-6 –mediated transcriptional activation. (A) N/N1003a cells, (B) ␣TN4-1 cells, (C) CHO cells, (D) T-antigen transformed 3T3 cells. E1b: E1b minimal promoter driving CAT; pKW10: six copies of Pax-6 consensus PD cloned in front of the E1b promoter driving CAT cotransfected with the parental pKW10 expression vector; Pax-6: six copies of Pax-6 consensus PD cloned in front of the E1b promoter driving CAT cotransfected with the pKW10 vector driving expression of the fulllength, consensus Pax-6 protein; PD ⫹ HD: six copies of Pax-6 consensus PD cloned in front of the E1b promoter driving CAT cotransfected with the pKW10 expression vector driving expression of a truncated Pax-6 protein (aa 1–286) containing the PD and HD; PD: six copies of Pax-6 consensus PD cloned in front of the E1b promoter driving CAT cotransfected with the pKW10 expression vector driving expression of a truncated Pax-6 protein (aa 1–140) containing the PD.

also had small lenses, vacuolated cortical fibers, and a notable eosin-dense posterior nuclear cataract (Fig. 3C). The lens defects in the PD ⫹ HD and PD expressing transgenic mouse lines arose mostly during the postnatal period, because newborn mice of these strains have lenses that appear close to normal with only slight alterations in the morphology of fiber cells (data not shown). To confirm that the transgenes result in the expression of nuclear localized protein, immunofluorescence was performed on sections adjacent to those used for histology, by using a pan-specific PD antibody. Under the staining conditions used, the endogenous expression of Pax-6 protein (or any other PD-containing protein) was not detected in the lens (Fig. 3D). In contrast, immunofluorescence corresponding to PD expression was easily detected in the cortical fiber cell nuclei of mice harboring the PD ⫹ HD (Fig. 3E) or PD (Fig. 3F) transgenes. Notably, the PD ⫹ HD transgene was detected only in the nuclei of lens fiber cells, whereas the PD transgene was detected in the nuclei and faintly in the cytoplasm of the fiber cells.

Elevated Levels of Calcium and Mass Reduction in Lenses from ␣A-Crystallin–Truncated Pax-6 Transgenic Mice Because the lenses from transgenic mice appeared smaller than normal (Fig. 3), the total wet mass and the total mass of nonvolatile components (mostly protein, nucleic acid and lipid) of transgenic and normal mouse lenses (Table 1) was quantitated. The lenses of transgenic mice were approximately 50% of the wet mass of those from their nontransgenic littermates. In addition, the lenses of transgenic mice were relatively more hydrated than normal lenses, resulting in the transgenic mouse lenses’ having only 23% of the dry mass of the normal mouse lens. Because cataractogenesis in rodents is often associated with elevated levels of total calcium,35,36 total levels of lens calcium were measured by atomic absorption spectroscopy. The lenses from mice expressing the PD and PD ⫹ HD transgenes both have significantly higher total calcium levels than normal mice (Table 1).

468

Duncan et al.


Lens Protein Profiles for Mice Transgenic for ␣A-Crystallin–Truncated Pax-6 Transgenes

FIGURE 2. Gross morphology of lenses from mice transgenic for ␣Acrystallin Pax-6 transgenes. (A) Lens from nontransgenic littermate of the mouse whose lens is shown in (B). The faint ring seen around the periphery is an internal reflection of the annular light used for illumination. (B) Lens from a mouse harboring an ␣A-crystallin–PD ⫹ HD transgene showing a dense nuclear– cortical cataract. (C) Lens from a mouse harboring an ␣A-crystallin–PD transgene showing a dense posterior polar cataract. The opacity is out of focus because the photograph was taken from the anterior surface. (D) The same lens shown in (C) photographed from the posterior surface.

Because the reduced dry mass of the transgenic lens and the presence of Pax-6 – binding sites in crystallin promoters37 suggested that the transgenes could be altering crystallin synthesis, the total crystallin profile of adult lenses was determined by SDS-PAGE (Fig. 4A). The transgenic mice expressing the PD ⫹ HD construct have an additional band (arrow) that was not detected in lenses from either PD transgenic mice or nontransgenic littermates. However, the significance of this result is uncertain, because the amount of total soluble protein extractable from the transgenic lens was greatly reduced (data not shown). Western blot analysis was then performed to investigate further the relative amount of transgene expression compared with that of wild-type Pax-6. Consistent with previous observations in the chicken lens,17 wild-type Pax-6 protein was not detectable in either transgenic or nontransgenic adult mouse lenses by western blot analysis. Both lines of ␣A-crystallin–PD transgenic mice expressed easily detectable amounts of truncated PD protein, whereas neither line of ␣A–PD ⫹ HD transgenics exhibited levels of transgene expression detectable by this method.

Scanning Electron Microscopy Next, the shape, spacing, and organization of lens fiber cells was examined by scanning electron microscopy of microdissected lenses. In the normal lens, fiber cells were organized

FIGURE 3. Histologic analysis of lens from mice transgenic for ␣A-crystallin Pax-6 transgenes. (A, B, and C) Paraffin-embedded lens sections from 12-week-old mice stained with hematoxylin and eosin. (A) Lens from a nontransgenic littermate of the mouse shown in (C). Note the evenly stained cytoplasm. (B) Lens from a mouse transgenic for ␣A-crystallin–PD ⫹ HD. Note the overall reduction in size, the patchy staining of the lens nucleus with eosin, and the abnormal cortical fiber cell structure. (C) Lens from a mouse transgenic for ␣A-crystallin–PD. Note the densely stained posterior polar cataract and the presence of vacuoles in the cortical fibers. (D, E, and F) Paraffin-embedded lens sections from 12-week-old mice stained with a pan-specific PD antibody. (D) Section through a wild-type mouse lens adjacent to that in (A). Note that under the staining conditions used, endogenous Pax proteins were not detected in the lens. (E) Section through a ␣A-crystallin–PD ⫹ HD mouse lens adjacent to the one shown in (B). Note the localization of the PD ⫹ HD protein in the nuclei of lens fiber cells. (F) Section through a ␣A-crystallin–PD mouse lens adjacent to the one shown in (C). Note the localization of the PD protein in the nuclei and cytoplasm of lens fiber cells. Magnification, (A, B, and C) ⫻50; (D, E, and F) ⫻630. e, epithelium.



469

TABLE 1. Lens Weight, Dry Mass, and Total Calcium Construct

Lens Weight* (g)

n

Dry Weight* (%)

Reduction in Dry Mass (%)

Nontransgenic ␣A/PD ⫹ HD ␣A/PD Sey⬍Dey⬎

0.011 ⫾ 0.003 0.005 ⫾ 0.002 0.006 ⫾ 0.002 0.006 ⫾ 0.001

8 4 6 2

30 ⫾ 6 17 ⫾ 9 13 ⫾ 9 19 ⫾ 4

— 77 77 60

Caⴙⴙ/Wet Weight*† 0.06 ⫾ 0.01 1⫾1 1.2 ⫾ 0.7 1.8 ⫾ 0.7

The PD ⫹ HD transgenic lens is 44% of the wet mass of normal (P ⱕ 0.0003), is more hydrated than normal (P ⱕ 0.005), and has 16-fold higher than normal Ca⫹⫹ levels (P ⱕ 0.0001). The PD transgenic lens is 53% of the wet mass of normal (P ⱕ 0.0008), is more hydrated than normal (P ⱕ 0.0002), and has 20-fold higher than normal Ca⫹⫹ levels (P ⱕ 0.0006). The Sey⬍Dey⬎ lens is 53% of the wet mass of normal (P ⱕ 0.05), is more hydrated than normal (P ⱕ 0.05), and has 30-fold higher than normal Ca⫹⫹ levels (P ⱕ 0.0001). * Data are expressed as mean ⫾ SD. † Micrograms Ca⫹⫹ per milligram wet weight.

into concentric shells that wrap around 180° of the lens forming sutures at the intersection of their tips (Fig. 5A). These cells interdigitated with their neighbors laterally, forming the classic ball-and-socket junctions of lens fiber cells.38 In contrast, transgenic mice expressing the PD ⫹ HD of Pax-6 in their lenses had poorly organized shells of fibers and did not form well-organized sutures (Fig. 5B). The fiber cells appeared swollen with attenuated ball-and-socket junctions (Figs. 5E, 5H). In transgenic mice expressing the PD of Pax-6, the superficial fiber cells were only slightly swollen and formed an anterior suture (data not shown) and ball-and-socket junctions (Figs. 5C, 5F). However, on the posterior surface, the suture did not close over the posterior nuclear cataract, which consisted largely of cell debris (Fig. 5I).

Morphology of the Lens in Small Eye Mice Heterozygous mice carrying the small eye⬍Dey⬎ allele, a large deletion mutation encompassing the Pax-6 gene as well as the Wilms’ tumor locus,12 have been reported to have cataracts and microphthalmia first notable at eye lid opening.39 Because a dominant effect of a truncated protein does not explain this cataract, we re-examined the adult phenotype of heterozygous small eye⬍Dey⬎ mice. Although small eye⬍Dey⬎ mice have cloudy eyes that superficially appear cataractous, dissection and histology of these eyes reveal that the clouding of the visual axis is predomi-

nately caused by corneal abnormalities (M. K. Duncan, unpublished data, 1997). Generally, the lenses were clear but small in young adult mice (data not shown). In aged mice of 8 months to 1 year of age, we have observed one cortical cataract in 12 lenses surveyed (Fig. 5A). These lenses were investigated morphologically by both light (Fig. 5B) and scanning electron (Fig. 5C) microscopy. Generally, the lenses were relatively normal morphologically (Fig. 5B), with few obvious abnormalities except for a malformed posterior suture (Fig. 5C). There is significant variability in the eye of the small eye⬍Dey⬎ phenotype: Approximately 10% of the eyes examined were microphthalmic with severely disrupted morphology (data not shown). However, nongenetic developmental processes clearly play a role in this phenomenon, because, typically, these mice displayed unilateral microphthalmia. Although the small eye⬍Dey⬎ lens was relatively normal morphologically, it was smaller and more hydrated, with higher levels of total calcium than normal (Table 1). Because the small eye⬍Dey⬎ mutation results in the deletion of a number of genes besides Pax-6,12 the phenotype of small eye⬍Neu⬎, a mutation that should result in a truncated Pax-6 protein containing the PD and HD but without the C-terminal 115 amino acids4 was also investigated. When dissected, the lenses of small eye⬍Neu⬎ mice were small and mostly clear, similar to those of small eye⬍Dey⬎ mice (Fig. 5D); however, small anterior polar cataracts were occasionally seen (Fig. 5E). Scanning electron microscopy suggests that the fiber cell morphology of small eye⬍Neu⬎ mice was relatively normal (Fig. 5F).

DISCUSSION

FIGURE 4. Characterization of the lens protein profile of mice transgenic for ␣A-crystallin Pax-6 transgenes. (A) SDS-PAGE analysis of lens proteins from wild-type, ␣A-crystallin–PD, and ␣A-crystallin–PD ⫹ HD mice. (B) Western blot analysis of Pax-6 expression in lenses from wild-type, ␣A-crystallin–PD, and ␣A-crystallin–PD ⫹ HD mice and the Pax-6 expressing cell line, ␣TN4-1.19 h1 and h2, littermates transgenic for the PD ⫹ HD transgene; p1 and p1, littermates transgenic for the PD transgene; w, wild-type littermates of the PD transgenic mice; t, ␣TN4-1 cell extract.

Pax-6 is clearly important for the early induction of the eye and nose40 and the development of the endocrine pancreas8 and brain.7 Later, the developing eye appears to be extremely sensitive to the amount of Pax-6 expressed, because persons heterozygous for Pax-6 mutations4 and transgenic mice overexpressing Pax-6 both have iris and corneal defects.14 It is well accepted that most types of Pax-6 mutations (including nulls, point mutants, and truncations) give rise to iris and anterior chamber defects of similar severity.1 These clinical data have supported the idea that aniridia is caused by a haploinsufficiency of Pax-6 function, and the spectrum of eye defects seen in these patients arise from a combination of developmental plasticity and genetic background.1

470

Duncan et al.


FIGURE 5. Scanning electron microscopy of lens fiber cell morphology from 12-week-old mice transgenic for ␣A-crystallin Pax-6 transgenes. (A, D, and G) Lens from a nontransgenic 12-week-old mouse. (B, E, and H) Lens from a mouse transgenic for the ␣A-crystallin–PD ⫹ HD transgene. (A) The normal lens has regularly organized shells of fiber cells. (B) The PD ⫹ HD transgenic lens is smaller and does not have well-organized fiber cell shells. (C) The PD transgenic lens is smaller and has a prominent zone of disorganization at the posterior pole. (D) The normal lens has fiber cells aligned in parallel and highly interdigitated through ball-and-socket junctions. (E) The cortical fiber cells of the PD ⫹ HD transgenic lens appear swollen. The ball-and-socket joints are visible but are extremely abnormal morphologically. (F) The cortical fiber cells of the PD transgenic lens are slightly abnormal. They have ball-and-socket junctions but the interdigitations are elongated. (G) The lens fiber cells of a nontransgenic littermate (micrograph is from near the posterior pole) have a normal pattern of fiber cell interdigitations. (H) The lens fiber cells of PD ⫹ HD transgenic mice deeper in the lens than shown in (E) are visibly distorted and wrinkled but still appear to be interdigitated (arrow). (I) The posterior polar cataract of PD transgenic mice has no structures that have fiber cell morphology and appears to consist entirely of cell debris. Magnification, (A, B, and C) ⫻50; (D, E, and H) ⫻3000; (F, G, and I) ⫻5000.

It is well established that aniridia often also involves cataract. Recently, however, it has been noted that individuals with Pax-6 mutations that prematurely truncate the PST domain are especially prone to cataract.23 In fact, a family has been reported in which such a mutation causes cataracts in the absence of overt aniridia.5 Additionally, it has been shown that truncated forms of Pax-6 bind DNA with a higher affinity than wild-type Pax-6 and are capable of blocking the ability of wild-type Pax-6 to transactivate an artificial promoter containing Pax-6 PD-binding sites.18 In the present

study, we showed that truncated Pax-6 proteins consisting of either the PD ⫹ HD or PD alone could prevent wild-type Pax-6 from activating an artificial promoter consisting of different Pax-6 PD-binding sites upstream of the E1b minimal promoter in lens cell lines. Moreover, truncated Pax-6 protein consisting of PD ⫹ HD blocked the activation of this promoter by wild-type Pax-6 in nonlens cells but, curiously, truncated Pax-6 consisting of only the PD did not repress Pax-6 activation of the artificial promoter in transfected nonlens cells. The biochemical basis for this cell line spec-



471

FIGURE 6. A re-examination of the lens phenotype of heterozygous small eye mice. (A) Gross morphology of the small eye⬍Dey⬎ lens showing the only opacity observed of 12 lenses examined. (B) Hematoxylin and eosin–stained section through the lens of a small eye⬍Dey⬎ mouse. Note the relatively normal lens architecture. (C) Scanning electron micrograph of a microdissected lens from a small eye⬍Dey⬎ mouse. Note the relatively normal fiber cell morphology and the posterior sutural defect. (D) Gross morphology of the small eye⬍Neu⬎ lens showing the typical clear phenotype. (E) Hematoxylin and eosin–stained section through the lens of a small eye⬍Neu⬎ mouse. Note the small anterior subcapsular–polar cataract. This was an atypical finding; the morphology of the lens epithelium was mostly normal. (F) Scanning electron micrograph of a microdissected lens from a small eye⬍Neu⬎ mouse. Note the nearly normal fiber cell morphology. e, epithelium; f, fibers; sd, sutural defect; apc, anterior polar cataract; cc, cortical cataract.

ificity is unclear but may suggest that different truncations can affect Pax-6 function differently in different tissues. Because the present transfection and previous biochemical data17,18 suggest that both PD ⫹ HD and PD can interfere with the activity of wild-type Pax-6 in cultured lens cells, we tested the ability of these truncated proteins to produce a lens phenotype in vivo by creating transgenic mice that overexpressed these truncated forms of Pax-6 in the lens under the control of the ␣A-crystallin promoter. Both truncations produced transgenic mice in which cataracts developed, albeit to different extents. These cataracts do not appear to be due to defects in fiber cell elongation or denucleation, because these mice have relatively normal lenses at birth. Instead, the fibers of the cataractous lenses are swollen, more hydrated than normal, and have elevated levels of total tissue calcium, all of which are typical of cataractogenesis.41,42 Although the result of the presence of truncated Pax-6 molecules in the lens is cataract, the initiating event for these cataracts is not clear. In vitro experiments suggest that truncated Pax-6 molecules bind to Pax-6 – binding sites found in the promoters of important lens proteins37 and block their transcriptional activity.18 In the simplest interpretation of this scenario, the cataracts are caused by the absence of one or more gene products from the lens whose expression is normally dependent on Pax-6 in lens fiber cells. This interpretation is complicated by the fact that the mouse ␣A-crystallin promoter used to drive transgene expression is dependent on Pax-6 for its function.19 It is possible that expression of PD ⫹ HD or PD may be repressing the transgene promoter itself and that transgene expression may be pulsatile. It is also possible that the truncated Pax-6 proteins cause cataracts by disrupting

the DNA interactions of non–Pax-6 transcription factors, the DNA-binding sites of which overlap with a Pax-6 site. The overlap of Pax-6 – binding sites with those of other factors has been shown previously to result in additive transcriptional activation in some cases43 and transcriptional repression in others.17,44 Western blot analysis has demonstrated that a large excess of PD protein is present in the transgenic lens, and competition with endogenous factors for DNA-binding sites is therefore a plausible scenario. Whereas much lower levels of the PD ⫹ HD are expressed, it is still possible that it blocks the binding of positively acting transcription factors to DNA because a similarly truncated molecule was demonstrated to have four times the affinity for DNA as wild-type Pax-6.18 Alternatively, it is possible that the Pax-6 transgenes upregulate the expression of wild-type Pax-6 in lens fiber cells, because expression of the Pax-6 gene may be autoregulated by Pax-6 protein.45 However, this possibility is relatively unlikely, because Pax-6 is normally not transcribed in lens fiber cells46 and the truncated Pax-6 proteins used in this study have no transcriptional activation domain.5,18 Unfortunately, the direct effect of truncated Pax-6 proteins on crystallin expression in the transgenic lens is difficult to ascertain. SDS-PAGE analysis of transgenic and wild-type lenses strongly suggests that the relative amount of each crystallin polypeptide made is unchanged, even though the dry mass of transgenic lenses is 77% lower than in wild type lenses. This may suggest that the cataracts observed in the Pax-6 truncation– expressing transgenic mice were due to the transgenes’ directly affecting the transcription of all crystallin genes equally. Alternatively, the truncated Pax-6 molecules may have

472

Duncan et al.

affected the expression of other (as yet unidentified) proteins that control the amount of crystallin proteins synthesized by the lens. For instance, because transgenic lenses have elevated levels of calcium, it is possible that the truncated Pax-6 molecules may have altered the expression of proteins important for calcium homeostasis. This would result in alterations in the many signal transduction pathways regulated by calcium,47 which could explain the observed alterations in lens biology. Finally, the global repression of crystallin gene expression in the lens may be a result of the truncated Pax-6 proteins’ binding to other cellular factors (such as retinoblastoma or TATA box– binding protein48), effectively removing their functions from the cellular pool. Cataracts are a typical feature of aniridia, with anterior subcapsular and polar opacities being most common.23 That Pax-6 mRNA is expressed principally in the cuboid epithelium on the anterior surface of the developing lens and the proliferative zone located at the lens equator9,22,46 suggests that Pax-6 helps prevent the epithelial to mesenchyme transitions49 that result in anterior polar cataract. However, cortical and posterior polar opacities have also been commonly reported in aniridia.2,50 Unfortunately, even though congenital cataracts have been reported to be more prevalent in aniridia in patients who harbor truncation mutations in the PST domain,23 the clinical description of the cataracts was not reported. The Sey⬍Neu⬎ mouse, which harbors a truncation mutation in the PST domain,4 as well as the Sey⬍Dey⬎ mouse, which harbors a complete deletion of the Pax-6 gene,12 do not exactly recapitulate the phenotype of the human aniridic eye. Most notably, the lenses of these mice generally remain clear throughout life, with only occasional cataractous changes seen in the anterior epithelium. It is likely that the differences in the lens phenotype between mice and humans harboring Pax-6 truncation mutations reflect either differences in the time that lens cells are exposed to the truncated proteins or differences in the duration of the life span of individual lens fiber cells. In the present study, we successfully generated both cortical and posterior polar lens opacities by ectopically expressing truncated forms of Pax-6 in lens fiber cells. Because some truncated Pax-6 proteins bind to DNA with a fourfold higher affinity than wild-type Pax-6 and can block the function of wild-type Pax-6 in transfection tests, it can be proposed that the cataractous changes seen in the lenses of patients with aniridia harboring truncation mutations of Pax-6 could be due to a dominant–negative effect. However, it is still possible that the predominant cause of aniridiaassociated cataract is either haploinsufficiency of Pax-6 in lens cells or a secondary effect caused by the other eye diseases seen in these patients. A definitive answer to this question awaits better clinical characterization of aniridiaassociated cataract in patients whose underlying Pax-6 mutation has been determined.

CONCLUSIONS Truncated Pax-6 molecules consisting of either the PD alone or the PD ⫹ HD can block the ability of Pax-6 to activate transcription through binding to PD-binding sites consistent with previous observations.18 Further, these potentially dominant– negative forms of Pax-6 can disrupt normal lens anatomy when

IOVS, February 2000, Vol. 41, No. 2 overexpressed in transgenic mice. Thus, it is possible that some of the phenotypic variability seen in aniridia is a result of a dominant–negative effect by the abnormal allele as well as a haploinsufficiency of wild-type Pax-6.

Acknowledgments The authors thank Larry David for numerous helpful discussions; and Susan Dickinson, Steven Lee, and Eric Wawrousek of the National Eye Institute transgenic mouse facility; Robert Nardone and Kirk Czymmek of the University of Delaware core molecular biology facility; Robert Ford of the Department of Plant and Soil Sciences, University of Delaware; and Robert Wieland of the Department of Material Science, University of Delaware, for technical support.

References 1. Churchill A, Booth A. Genetics of aniridia and anterior segment dysgenesis. Br J Ophthalmol. 1996;80:669 – 673. 2. Nelson LB, Spaeth GL, Nowinski TS, Margo CE, Jackson L. Aniridia: a review. Surv Ophthalmol. 1984;28:621– 642. 3. Glaser T, Walton DS, Cai J, et al. Pax6 gene mutations in aniridia: molecular genetics of ocular disease. New York:Wiley–Liss; 1995: 51– 82. 4. Hill RE, Favor J, Hogan BL, et al. Mouse small eye results from mutations in a paired-like homeobox-containing gene. Nature. 1991;354:522–525. 5. Glaser T, Jepeal L, Edwards JG, et al. Pax6 gene dosage effect in a family with congenital cataracts, aniridia, anophthalmia and central nervous system defects. Nat Genet. 1994;7:463– 471. 6. Grindley JC, Davidson DR, Hill RE. The role of Pax-6 in eye and nasal development. Development. 1995;121:1433–1442. 7. Stoykova A, Gotz M, Gruss P, Price J. Pax6-dependent regulation of adhesive patterning, R-cadherin expression and boundary formation in the developing forebrain. Development. 1997;124:3765– 3777. 8. Onge LS, Sosa–Pineda B, Chowdhury K, Mansouri A, Gruss P. Pax6 is required for differentiation of glucagon-producing alpha-cells in the mouse pancreas. Nature. 1997;387:406 – 409. 9. Koroma B, Yang J, Sundin O. The Pax-6 homeobox gene is expressed throughout the corneal and conjunctival epithelia. Invest Ophthalmol Vis Sci. 1997;38:108 –120. 10. Jun S, Desplan C. Cooperative interactions between paired domain and homeodomain. Development. 1996;122:2639 –2650. 11. Epstein JA, Glaser T, Cai J, et al. Two independent and interactive DNA-binding subdomains of the Pax6 paired domain are regulated by alternative splicing. Genes Dev. 1994;8:2022–2034. 12. Glaser T, Lane J, Housman D. A mouse model of the aniridia-Wilm’s tumor deletion syndrome. Science. 1990;250:823– 827. 13. Azuma N, Hotta Y, Tanaka H, Yamada M. Missense mutations in the Pax6 gene in aniridia. Invest Ophthalmol Vis Sci. 1998;39:2524 – 2528. 14. Schedl A, Ross A, Lee M, et al. Influence of Pax6 gene dosage on development: overexpression causes severe eye abnormalities. Cell. 1996;86:71– 82. 15. Hanson I, Churchill A, Love J, et al. Missense mutations in the most ancient residues of the PAX6 paired domain underlie a spectrum of human congenital eye malformations. Hum Mol Genet. 1999;8: 165–172. 16. Martha A, Strong LC, Ferrell RE, Saunders GF. Three novel aniridia mutations in the human PAX6 gene. Hum Mutat. 1995;6:44 – 49. 17. Duncan MK, HaynesII JI, Cvekl A, Piatigorsky J. Dual roles for Pax-6: a transcriptional repressor of lens fiber-cell specific ␤-crystallin genes. Mol Cell Biol. 1998;18:5579 –5586. 18. Singh S, Tang HK, Lee J-Y, Saunders GF. Truncation mutations in the transactivation region of Pax6 result in dominant–negative mutants. J Biol Chem. 1998;273:21531–21541. 19. Cvekl A, Kashanchi F, Sax CM, Brady JN, Piatigorsky J. Transcriptional regulation of the mouse alpha A-crystallin gene: activation dependent on a cyclic AMP-responsive element (DE1/CRE) and a Pax-6-binding site. Mol Cell Biol. 1995;15:653– 660.



20. Gopal–Srivastava R, Cvekl A, Piatigorsky J. Pax-6 and alphaB-crystallin/small heat shock protein gene regulation in the murine lens: interaction with the lens-specific regions, LSR1 and LSR2. J Biol. Chem. 1996;271:23029 –23036. 21. Cvekl A, Sax CM, Li X, McDermott JB, Piatigorsky J. Pax-6 and lens-specific transcription of the chicken delta 1-crystallin gene. Proc Natl Acad Sci USA. 1995;92:4681– 4685. 22. Richardson J, Cvekl A, Wistow G. Pax-6 is essential for lens-specific expression of zeta-crystallin. Proc Natl Acad Sci USA. 1995;92: 4676 – 4680. 23. Gupta SK, Becker ID, Tremblay F, Guernsey DL, Neumann PE. Genotype/phenotype correlations in aniridia. Am J Ophthalmol. 1998;126:203–210. 24. Reddan JR, Chepelinsky AB, Dziedzic DC, Piatigorsky J, Goldenberg EM. Retention of lens specificity in long-term cultures of diploid rabbit lens epithelial cells. Differentiation. 1986;33:168 – 174. 25. Yamada T, Nakamura T, Westphal H, Russell P. Synthesis of alphacrystallin by a cell line derived from the lens of a transgenic animal. Curr Eye Res. 1990;9:31–37. 26. O’Toole TE, Katagiri Y, Faull RJ, et al. Integrin cytoplasmic domains mediate inside-out signal transduction. J Cell Biol. 1994;124: 1047–1059. 27. Aaronson SA, Todaro GT. Development of 3T3-like lines from Balb/c mouse embryo cultures: transformation susceptibility to SV40. J Cell Physiol. 1968;72:141–148. 28. Salamon C, Chervenak M, Piatigorsky J, Sax CM. The mouse transketolase (TKT) gene: cloning, characterization, and functional promoter analysis. Genomics. 1998;48:209 –220. 29. Overbeek PA, Chepelinsky AB, Khillan JS, Piatigorsky J, Westphal H. Lens-specific expression and developmental regulation of the bacterial chloramphenicol acetyltransferase gene driven by the murine alpha A-crystallin promoter in transgenic mice. Proc Natl Acad Sci USA. 1985;82:7815–7819. 30. Duncan MK, Roth HJ, Thompson M, Kantorow M, Piatigorsky J. Chicken ␤B1-crystalin: gene sequence and evidence for functional conservation of promoter activity between chicken and mouse. Biochim Biophys Acta. 1995;1261:68 –76. 31. Reneker LW, Overbeek PA. Lens-specific expression of PDGF-A alters lens growth and development. Dev Biol. 1996;180:554 –565. 32. Krausz E, Augusteyn RC, Quinlan RA, et al. Expression of Crystallins, Pax6, Filensin, CP49, MIP, and MP20 in lens-derived cell lines [see comments]. Invest Ophthalmol Vis Sci. 1996;37:2120 –2128. 33. Tang HK, Chang L-Y, Saunders GF. Functional analysis of paired box missense mutations in the Pax-6 gene. Hum Mol Genet. 1997;6:381–386. 34. Chepelinsky AB, Khillan JS, Mahon KA, et al. Crystallin genes: lens specificity of the murine alpha A-crystallin gene. Environ Health Perspect. 1987;75:17–24. 35. Hightower KR, David LL, Shearer TR. Regional distribution of free

36.

37. 38. 39.

40. 41.

42.

43.

44.

45.

46.

47. 48.

49.

50.

473

calcium in selenite cataract: relation to calpain II. Invest Ophthalmol Vis Sci. 1987;28:1702–1706. David LL, Azuma M, Shearer TR. Cataract and the acceleration of calpain-induced ␤-crystallin insolubilization occuring during normal maturation of rat lens. Invest Ophthalmol Vis Sci. 1994;35: 785–793. Cvekl A, Piatigorsky J. Lens development and crystallin gene expression: many roles for Pax-6. Bioessays. 1996;18:621– 630. Kuszak JR. The ultrastructure of epithelial and fiber cells in the crystallin lens. Int Rev Cytol. 1995;163:305–350. Theiler K, Varnum DS, Stevens LC. Development of Dickie’s small eye, a mutation in the house mouse. Anat Embryol. 1978;155:81– 86. Hill RE, Hanson IM. Molecular genetics of the Pax gene family. Curr Opin Cell Biol. 1992;4:967–972. Azuma M, David LL, Shearer TR. Hydration and elevated calcium alone do not produce xylose nuclear cataract: role of proteolysis by calpain. Ophthalmic Res. 1992;24:8 –14. Kuszak JR, Deutsch TA, Brown HG. Biochemistry of the crystallin lens: anatomy of aged and senile cataractous lenses. In: Albert DM, Jakobiec FA, eds. Clinical Practice: Principles and Practice of Ophthalmology. Vol 1. Philadelphia: WB Saunders; 1994:564 –575. Gopal–Srivastava R, Cvekl A, Piatigorsky J. Involvement of retinoic acid/retinoid receptors in the regulation of murine alphaB-crystallin/small heat shock protein gene expression in the lens. J Biol Chem. 1998;273:17954 –17961. Plaza S, Grevin D, MacLeod K, Stehelin D, Saule S. Pax-QNR/Pax-6, a paired- and homeobox-containing protein, recognizes ETS binding sites and can alter the transactivating properties of ETS transcription factors. Gene Expression. 1994;4:43–52. Okladnova O, Syagailo YV, Mossner R, Riederer P, Lesch K–P. Regulation of Pax-6 gene transcription: alternate promoter usage in human brain. Mol Brain Res. 1998;60:177–192. Li H-S, Yang J-M, Jacobson RD, Pasko D, Sundin O. Pax-6 is first expressed in a region of ectoderm anterior to the early neural plate: implications for stepwise determination of the lens. Dev Biol. 1994;162:181–194. Hardingham GE, Bading H. Nuclear calcium: a key regulator of gene expression. Biometals. 1998;11:345–358. Cvekl A, Kashanchi F, Brady JN, Piatigorsky J. Pax-6 interactions with TATA-box-binding protein and retinoblastoma protein. Invest Ophthalmol Vis Sci. 1999;10:1343–1350. Schmitt–Graff A, Pau H, Piper HM, Skalli O, Gabbiani G. Appearance of alpha-smooth muscle actin in human eye lens cells of anterior capsular cataract and in cultured bovine lens-forming cells. Differentiation. 1990;43:115–122. Andersen SR, Geertinger P, Larsen H–W, et al. Aniridia, cataract, and gonadoblastoma in a mentally retarded girl with deletion of chromosome 11. Ophthalmologica 1978;176:171–177.