May 2002
Volume 43, Issue 5
Free
Biochemistry and Molecular Biology  |   May 2002
Foxe3 Haploinsufficiency in Mice: A Model for Peters’ Anomaly
Author Affiliations
  • Mattias Ormestad
    From the Department of Molecular Biology, The Lundberg Laboratory, and the
  • Åsa Blixt
    From the Department of Molecular Biology, The Lundberg Laboratory, and the
  • Amanda Churchill
    Department of Molecular Medicine and Ophthalmology, St. James’s Hospital, Leeds University, Leeds, United Kingdom;
  • Tommy Martinsson
    Department of Clinical Genetics, Sahlgrenska University Hospital, Gothenburg, Sweden.
  • Sven Enerbäck
    Department of Medical Biochemistry, Gothenburg University, Gothenburg, Sweden; the
  • Peter Carlsson
    From the Department of Molecular Biology, The Lundberg Laboratory, and the
Investigative Ophthalmology & Visual Science May 2002, Vol.43, 1350-1357. doi:
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Mattias Ormestad, Åsa Blixt, Amanda Churchill, Tommy Martinsson, Sven Enerbäck, Peter Carlsson; Foxe3 Haploinsufficiency in Mice: A Model for Peters’ Anomaly. Invest. Ophthalmol. Vis. Sci. 2002;43(5):1350-1357.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

purpose. To evaluate the importance in anterior segment dysgenesis of genetic variation in Foxe3, a gene encoding a forkhead transcription factor specifically expressed in the lens.

methods. The phenotype of mice heterozygous for a mutation in the DNA-binding domain of Foxe3 was examined from histologic sections, and DNA binding by the encoded protein was investigated by gel-shift assay. FOXE3 from human patients with Peters’ anomaly was PCR amplified and sequenced.

results. The dysgenetic lens (dyl) allele of Foxe3 was found to encode a protein unable to bind DNA. Approximately 40% of mice heterozygous for Foxe3 dyl have corneal and lenticular defects. The phenotype is variable but typically consists of the equivalent of Peters’ anomaly in humans, with central corneal opacity, keratolenticular adhesion, and, in some cases, anterior polar cataract. In a small cohort (n = 13) of patients with Peters’ anomaly, shown to be normal in the PAX6 locus, one individual was found to be heterozygous for a nonconservative missense mutation in FOXE3. The mutation, which does not occur in 116 chromosomes from a control population, substitutes leucine for arginine 90 at a highly conserved position in the forkhead domain.

conclusions. Haploinsufficiency of Foxe3 in a mouse model causes anterior segment dysgenesis similar to Peters’ anomaly. Although causality could not be shown in the human case, the presence of a rare, nonconservative substitution in FOXE3 of a patient with Peters’ anomaly is interesting, in light of the phenotypic similarities with the mutant mice.

Development of the anterior segment of the eye is a complex process that depends on multiple inductive events and coordinated interactions between cells of ectodermal, neuroectodermal, and neural crest origin. Disturbances in this process can cause glaucoma, cataract, leucoma, or other conditions that lead to vision impairment or blindness. Dominant effects of loss-of-function mutations—haploinsufficiency—are fairly common causes of congenital developmental malformations and frequently involve genes encoding transcription factors or signaling molecules. 1 Anterior segment morphogenesis appears to be particularly sensitive to deviations in expression levels of the regulatory genes on which it depends. Mutations in a number of transcription factor genes—all of which are involved in the control of developmental processes in other organs as well—cause congenital anterior segment malformations in the heterozygous state. The forkhead gene Foxc1 (formerly FREAC3 or FKHL7 in humans, Mf1 in mouse) is widely expressed in mesodermal- and neural crest-derived tissues during embryonic development. Homozygous Foxc1-null mutants have congenital hydrocephalus, 2 whereas heterozygosity for FOXC1 mutations are associated with congenital glaucoma, iridogoniodysgenesis, and Axenfeld-Rieger anomaly. 3 4 5 Null mutants of a related forkhead gene, Foxc2 (formerly Mfh1), have a similar heterozygous phenotype restricted to the anterior segment of the eye, 6 although homozygotes have pleiotropic defects that are distinct from those of Foxc1 mutants. 7 8 In humans, FOXC2 haploinsufficiency is associated with lymphedema-distichiasis syndrome. 9 Iridogoniodysgenesis syndrome and Axenfeld-Rieger syndrome are caused by haploinsufficiency of the homeobox gene PITX2 (formerly RIEG). 10 PAX6 mutations give rise to variable anterior segment malformations: aniridia and Peters’ anomaly in humans and similar lens, iris, and corneal defects in mouse. 11 12 13 14 15 The anterior segment of the eye exhibits the most severe phenotype in Pax6 heterozygotes, although subtle brain malformations can also be detected 16 ; homozygous mutants have no eyes and have gross defects in brain and nasal development. 17 18 19 Both Pax6 and FOXC1 cause eye malformations, not only when the gene dosage is decreased, but also when it is increased, 20 21 which further emphasizes the importance of precise control of transcription factor expression levels in eye development. The phenotypical similarity between mouse and humans in these examples also illustrates the usefulness of mouse models for human genetic eye disorders. 
Foxe3 (formerly FREAC8) encodes a forkhead transcription factor that is expressed in the lens anlage during eye formation and in lens epithelium thereafter. 22 23 24 The mouse mutant dysgenetic lens (dyl) is homozygous for mutations in Foxe3 that cause two amino acid substitutions in the DNA-binding domain. 23 24 Dyl mice have serious malformations of the anterior segment, with fusion of lens, iris, and cornea; corneal dysplasia; lens and iris hypoplasia; and severe cataract. 25 Analysis of the mutant phenotype showed that Foxe3 is essential for proliferation, prevention of premature differentiation, and protection against apoptosis in the lens epithelium. 23 Semina et al. 26 recently reported a mutation in FOXE3 in a patient with anterior segment dysgenesis and cataract. Based on the severe malformations in the homozygous dyl mutant, Sanyal and Hawkins 25 suggested that dyl may be a candidate gene for Peters’ anomaly in humans. Occasionally, this condition is caused by mutations in PAX6, 12 PITX2, 27 or CYP1B1, 28 but in the majority of cases the genetic basis remains unknown. 29 In the current study, we investigated the phenotype of mice heterozygous for the Foxe3 dyl mutation. Although dyl is described as a recessive mutation, 25 we showed that Foxe3 haploinsufficiency has a penetrance of approximately 40% and gives rise to a variable phenotype that includes the equivalent of Peters’ anomaly in humans. Finally, we sequenced FOXE3 in a small cohort of patients with Peters’ anomaly and found an individual with a missense mutation in the forkhead domain. 
Methods
Cloning and Sequencing the FOXE3 cDNA
Two cDNA libraries made from human lens epithelium (kindly provided by Toshimichi Shinohara, Harvard Medical School, Boston, MA, and Allan R. Shepard, Mayo Clinic, Rochester, NY) were screened with the original FOXE3 clone (published under the name FREAC8). 22 Four overlapping cDNA clones were isolated and sequenced in their entirety on a CEQ2000 sequencer (Beckman Instruments, Carlsbad, CA). 
Radiation Hybrid Mapping
The Stanford G3 radiation hybrid panel (Research Genetics, Huntsville, AL) was screened by PCR with the primers 5′-AAGTGTCCTCAGGGCGTGAAG and 5′-TCAGACTCCTGGGTTCATGACTTA, which amplify part of the 3′ untranslated region of FOXE3. Positive and negative signals were scored, and the results were electronically submitted to the Stanford Human Genome Center (Stanford University, Stanford, CA). The primers were also used for analysis of a single-chromosome panel (i.e., the National Institute of General Medical Sciences [NIGMS] Human/Rodent Somatic Cell Hybrid Mapping Panel 2, derived from Coriell Cell Repositories, Camden, NJ). 
Protein Expression and Gel-Shift Assay
The forkhead boxes of Foxe3 (encoding amino acids 60-175) from wild-type and dyl mice and the corresponding region of FOXE3 from a patient with Peters’ anomaly (patient LAS) were amplified by PCR and cloned in a pGEX plasmid vector. T7 in vitro transcription, in vitro translation in reticulocyte lysate (Promega, Madison, WI) in the presence of [35S]-methionine (Amersham, Amersham, UK), and analysis of DNA binding by gel-shift assay were all performed as described. 30 Relative protein concentrations were estimated by autoradiography of 35S-labeled proteins run on SDS-PAGE. The gel-shift probe consisted of a 32P-labeled, double-stranded oligonucleotide, which contains a binding site for FOXC1 (FREAC3, probe “B” in Pierrou et al. 30 ). 
Mouse Strains
Wild-type mice used in this study were BALB/c obtained from Charles River, Inc. (Uppsala, Sweden). The dyl mutant, which arose spontaneously in the BALB/c strain, 25 was obtained from The Jackson Laboratory (Bar Harbor, ME) and was crossed with BALB/c to produce heterozygotes. PCR genotyping of the Foxe3 dyl allele was performed as described. 23 The use of mice in this study adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 
Histology
Eyes were fixed in 4% paraformaldehyde and processed for paraffin embedding using standard techniques. Sections (5 μm) were stained with hematoxylin and eosin. 
Sequencing of FOXE3 from Patients with Peters’ Anomaly
DNA was prepared from blood samples of affected individuals and the entire coding region of FOXE3 and 425 bp of flanking sequences were PCR amplified with the primers 5′-TGTCCATATAAAGCGGGTCGG and 5′-TATGGACTCACTGGAGGCGAG. The PCR products were sequenced directly with the same primers and with the internal primers 5′-GCG TCCAGTAGTTGCCCTTGC, 5′-GGAAGCCGCCCTACTCGTACA, and 5′-AGTTTGAGTCCAGGAGGCCACGAC. To verify the heterozygous nucleotide substitution of patient LAS, PCR products were cloned in a plasmid vector, and the sequence was determined from plasmid clones representing both alleles. Informed consent was obtained from all subjects involved in this study, which followed the tenets of the Declaration of Helsinki and was approved by the institutional human experimentation ethics committee. 
Results
The Human FOXE3 Gene
The original FOXE3 clone (published under the name FREAC8 22 ) was isolated from a human genomic library. To isolate the FOXE3 cDNA, we used the genomic clone to screen two cDNA libraries made from human lens epithelium. The sequence obtained from four overlapping clones contains an open reading frame of 957 nucleotides (nt), which predicts a protein of 319 amino acids (33.2 kDa) and an mRNA size of at least 1991 nt plus polyA-tail (Fig. 1a) . Comparison with genomic clones from mice 23 and humans 22 26 showed that FOXE3 has no introns. The average amino acid sequence similarity between the human and mouse protein is 82%, but the degree of conservation is unevenly distributed. The forkhead domains, which are responsible for DNA binding and span 112 amino acids, are 100% identical. Human FOXE3 is predicted to be approximately 10% longer than the mouse protein, because of a six-amino-acid extension in the amino terminus and four internal expansions of regions with low sequence complexity. 
We have previously localized FOXE3 by fluorescence in situ hybridization (FISH) to human chromosome 1p32 22 and the mouse homologue Foxe3 to the syntenic region 4C7. 23 To determine the position of the human locus more accurately, we screened the Stanford G3 radiation hybrid panel with PCR primers derived from the 3′ flanking region of FOXE3. Highly significant linkages were observed between FOXE3 and two chromosome 1p expressed-sequence tag (EST) markers. Both ESTs, SHGC-35625, human DD96 mRNA (lod score 8.29; estimated distance from FOXE3 14 centirad [cR]-10000), and SHGC-36697, human EST (lod score 6.64; 23 cR-10000), mapped to the 2.8 centimorgan (cM)-72 cR-10000 region between markers D1S477 and D1S451. Thus, the data indicate that FOXE3 is located between markers D1S477 and D1S451 (i.e., in the proximal region of the scaffold interval D1S2843-D1S417. This localization is in perfect agreement with the previous localization by FISH to 1p32. Screening of a single-chromosome hybrid panel confirmed the assignment to chromosome 1 (data not shown). 
In the draft of the human genome, 31 a contig that has been mapped to 1p32 contains sequences that match FOXE3. The coding sequence is, however, split into two regions approximately 20 kb apart. From comparison of genomic clones 22 26 and the mRNA sequence presented herein, it is clear that FOXE3 does not contain introns. Because there are no other candidate matches to FOXE3 in the draft sequence and the position of the disrupted sequence in the genome is in perfect agreement with our localization of FOXE3, the discrepancy is more likely to represent an assembly artifact in the draft sequence than the existence of a pseudogene. The Celera draft sequence 32 also has an imperfect match to FOXE3 on a contig from chromosome 1. The partial match spans the 3′ untranslated region (UTR) and the end of the coding sequence of FOXE3. The position of the match in the public draft sequence has been estimated to 46 Mb on chromosome 1 and the Celera Match to 38 Mb. 
DNA Binding by Foxe3 dyl
The dyl mouse mutant has defects in lens development, and we have previously shown that dyl mice have two missense mutations in the Foxe3 forkhead box, which encodes the DNA-binding domain. 23 24 Based on the lens-specific expression of Foxe3 and cosegregation of Foxe3 mutations with the dyl phenotype, we argued that mutations in Foxe3 are likely to cause the lens defects, and we predicted that the amino acid substitutions obliterate DNA binding. 23 To test this directly, we produced the Foxe3 forkhead domain from wild-type and dyl mutant mice by in vitro transcription and translation. The DNA-binding specificity of Foxe3 has not yet been determined, and the sequence of the optimal binding site thus remains unknown. Instead, we tested binding of Foxe3 to a consensus site for FOXC1 (FREAC3 or FKHL7), which has been shown to bind also to FOXD1 (FREAC4 30 33 ) and FOXF2 (FREAC2; Peter Carlsson, unpublished data, 1994). Wild-type Foxe3 binds this site very well in a gel-shift assay, but the dyl mutant protein was unable to produce a shifted complex (Fig. 2) , even though a higher concentration of protein was used (based on SDS-PAGE of 35S-labeled protein). This shows that the dyl amino acid substitutions in the Foxe3 protein destroy its sequence-specific binding to DNA. Because the activity of Foxe3 depends on its ability to bind DNA, dyl most likely represents a null allele. 
Phenotype of Foxe3 Heterozygotes
In the original publication of the mutant, dyl is described as an autosomal recessive trait. 25 With regard to the grave anterior segment malformations, which are readily identified by microphthalmia and severe cataract, the heritability is clearly recessive. However, when scoring the phenotype of progeny from F1 crosses in the cosegregation analysis, we noticed that Foxe3 dyl/+ heterozygotes frequently had uni- or bilateral corneal irregularities, often in combination with a minor central leucoma or cataract (Fig. 3b) . Because a malformation due to haploinsufficiency is much more likely to be of significance in morbidity in humans than homozygosity for a recessive trait, we examined the eyes of 360 offspring from Foxe3 dyl/+ × Foxe3 dyl/+ crosses at approximately 3 weeks of age. Macroscopic abnormalities (irregularities, depressions or blisters in the cornea, central leucoma or cataract) were observed in 39% of Foxe3 dyl/+ heterozygotes, but only in one wild-type animal. To investigate the morphologic basis of the observed malformations, histologic sections were examined from heterozygous eyes with macroscopic defects. The most striking malformation consisted of a central stalklike connection between lens and cornea (Figs. 4b 4e 4f 4j) . This connection distorted the curvature of the cornea as well as that of the anterior face of the lens. The stalk appeared to consist of epithelial cells continuous with the subcapsular lens epithelium. The core of the stalk contained lens fiber material that extended forward, sometimes into the corneal stroma. In homozygous dyl embryos, the lens vesicle never closes and does not separate from the surface ectoderm during lens formation. 25 The heterozygote phenotype described herein is most easily explained as a less severe form of the same defect. The presence of cells left over from lens formation in the corneal stroma creates an abnormal histology that is likely to contribute to the observed leucoma, whereas clusters of disorganized cells underneath the lens capsule (Fig. 4g) may cause the local cataract seen in some lenses. In some eyes, the anterior surface of the lens and the posterior surface of the cornea adhered directly to each other, without a connecting stalk (Fig. 4l) , and, in others, the connection was lost, but left marks on the surface of both lens and cornea. Another source of leucoma in these mice is corneal swelling associated with a lax, cavernous structure of the stroma. This anomaly does not affect the whole cornea. It is typically found around the connecting stalk or centrally in eyes without lens adhesion (Figs. 4e 4h 4k) and appears to represent a less severe and local version of the corneal dysplasia observed in homozygous dyl mutants. No histologic abnormalities were detected in sections of eyes from heterozygous animals that appeared normal when examined macroscopically. 
A Patient with Peters’ Anomaly with a Mutation in FOXE3
The anterior segment abnormalities observed in Foxe3 dyl/+ heterozygous mice closely resemble Peters’ anomaly. 34 35 Most cases of this congenital defect of the anterior chamber of the eye, with central adhesion between lens and cornea and central leucoma as frequent manifestations, are sporadic, and in only a few cases have mutations been reported. PAX6 mutations are usually associated with aniridia, but are occasionally found in Peters’ anomaly. 12 Doward et al. 27 described a PITX2 mutation in a child with Rieger syndrome, who exhibited some ocular features in one eye that overlap the spectrum of Peters’ anomaly. Mutations in CYP1B1 is a major cause of primary congenital glaucoma, 36 and mutations in this gene have been diagnosed in one case of Peters’ anomaly. 28 However, neither of these genes is mutated in most patients with Peters’ anomaly, and no other genes have yet been found to be involved. 29 We sequenced the FOXE3 gene from 13 individuals with diagnosis of Peters’ anomaly. Occurrence in nine was sporadic and may represent recessive inheritance or new dominant mutations and in four was familial (from four families). Of these four families, three showed dominant inheritance, and one (with consanguineous parents) exhibited a recessive pattern. The clinical details for these individuals have been described previously, and they were shown to be negative for mutations in the coding region of the PAX6 gene. 29 All patients in this study were born with bilateral congenital corneal opacities with or without cataracts. Apart from translationally silent polymorphisms, we found a FOXE3 mutation in a familial case of Peters’ anomaly (patient LAS) with eccentric corneal opacities and glaucoma but not cataract. 29 The sequence trace showed a mixed T/G peak, indicative of heterozygosity (Fig. 5c) . To verify the presence of two alleles, we cloned the PCR fragments and sequenced individual plasmid clones, which had either a T or a G in the affected position (Figs. 5b 5d) . The G→T mutation occurs in cDNA position 524 and causes an Arg90Leu substitution, located in the DNA-binding domain of FOXE3 (Figs. 1b ; 5b 5c 5d ). The mutant protein binds DNA when produced by in vitro translation and tested in a gel-shift assay (data not shown), although we have not performed a quantitative measurement of the affinity. It gives rise to a complex with slightly higher mobility than the normal protein, which may be caused by the decreased net charge from the Arg→Leu substitution, or an altered three-dimensional (3-D) structure. To exclude the possibility that the mutation represents a polymorphism in FOXE3, we sequenced the relevant region in 58 healthy control subjects with similar ethnic background (34 English and 24 Swedish). None of these had the G524T mutation. Semina et al. 26 screened the FOXE3 gene from 251 individuals—patients with ocular disorders and normal control subjects—for mutations and listed all polymorphisms detected. They found five single-nucleotide polymorphisms with rare allele frequencies of between 1% and 32%, all of which are translationally silent or cause conservative substitutions. The G524T mutation was not detected in this sample of more than 500 chromosomes. Therefore, based on our own and published data, we exclude the possibility that Arg90Leu represents a common polymorphism. 
Since the collection of blood samples several years ago, the proband (patient LAS) has died and, unfortunately, it has not been possible to obtain DNA from other members of this family to check whether the sequence change segregates with the disease. 
Discussion
Foxe3 dyl and DNA Binding
Foxe3 is essential for key events in anterior segment morphogenesis. In mice homozygous for the dyl allele, lens vesicles never close or separate from the presumptive cornea, with absence of anterior chamber and fusion of lens, iris, and cornea as a consequence. 25 The small, severely cataractous lens gradually vanishes, because of degeneration of the lens epithelium and absence of secondary lens fiber formation. 23 This, in turn, affects the cornea, leading to disruption of the normal stromal architecture. 
Based on the 3-D-structure of the forkhead domain and the high degree of evolutionary conservation of the affected amino acids, we predicted that the substitutions caused by the dyl mutations would impede DNA binding by the encoded protein. 23 From comparison with FOXC1 mutants, Saleem et al. 37 recently suggested that the substitutions do not influence DNA binding, but instead interfere with trans-activation by Foxe3. Herein, we provide evidence that that the amino acid substitutions caused by the dyl mutations indeed obliterate DNA binding, which also supports the notion that Foxe3 dyl is a null allele. 
Foxe3 Haploinsufficiency
Approximately 40% of mice with just a single functional copy of Foxe3 have uni- or bilateral eye defects. The phenotype of Foxe3 heterozygotes described herein corresponds to Peters’ anomaly in humans, which is characterized by central corneal leucoma, keratolenticular adhesion and, sometimes, lenticular opacities. 34 35 38 Foxe3 transcription is not autoregulated (similar level and distribution of mRNA are present in dyl/dyl and wild-type animals 23 ) and therefore heterozygotes are expected to produce half the normal concentration of functional protein. Because the dyl mutant protein does not bind DNA, it is unlikely to exert a dominant negative effect. Although other means of interfering with the function of the normal protein cannot be formally excluded, haploinsufficiency due to reduction in concentration of active protein is the simplest explanation for the observed defects in heterozygous mice. 
The corneal endothelium serves as a barrier against the aqueous humor and drives outflow of water from the stroma by pumping out ions. Injury or defects in the endothelial monolayer leads to edematous swelling of the stroma and corneal opacity, as in congenital hereditary endothelial dystrophy. 39 40 The corneal swelling in Foxe3 heterozygotes may be caused by a similar mechanism and may reflect a local imperfection in the endothelial layer. Because Foxe3 is not expressed in the cornea, the lens epithelium must influence corneal differentiation indirectly. Classic transplantation studies have implicated the anterior surface of the lens as the source of diffusable factors important in differentiation of cornea. 41 42 43 The dysplasia of corneal stroma in homozygous and, to a lesser extent, heterozygous dyl mutants suggests that Foxe3 is essential for the production of these factors. 
FOXE3 and Peters’ Anomaly in Humans
The missense mutation that we found in FOXE3 of a patient with Peters’ anomaly gives rise to a nonconservative amino acid substitution in the FOXE3 protein. Judged by comparison with forkhead proteins for which the 3-D structure has been solved, 44 45 the Arg90Leu substitution occurs at the junction between helix 1 and 2 in the helix-turn-helix motif of the DNA-binding domain, but it is clear from our gel-shift experiment that the mutant protein is still capable of binding DNA. Of 84 forkhead sequences in the SwissProt database (provided by the Swiss Institute of Bioinformatics, Geneva, Switzerland; available at http://www.expasy.org at no charge to academic users), 71 have lysine or arginine in the position corresponding to Arg90, but none has a hydrophobic residue. This indicates that the positive charge has been conserved by evolution and has an important function. The turn between helix 1 and 2 protrudes from the face of the forkhead domain, opposite the surface that makes contact with DNA, 44 and residues located there are in an ideal position to interact with other proteins. Possible candidates include proteins involved in nuclear transport. These bind motifs with a positive charge, and the amino terminal part of the forkhead domain has been shown to be important in nuclear localization. 46 47  
In a recent paper, Semina et al. 26 reported identification of a heterozygous FOXE3 mutation in a patient with congenital cataract and posterior embryotoxon. A frameshift close to the carboxyl terminus replaces the last five amino acids of FOXE3 and extends the coding sequence with other residues. They conclude that DNA binding is not affected and speculate that trans-activation may be hampered. The mutant was identified in a sample of 161 patients with anterior segment anomalies, 28 of which were affected with congenital cataracts, anterior segment dysgenesis with cataracts, or Peters’ anomaly. The absence of mutations in 12 of the 13 patients reported in the current study and in the cases investigated by Semina et al. shows that mutations in FOXE3 are not a major cause of Peters’ anomaly. Instead it supports the notion that this is a genetically heterogenous condition. The high proportion of sporadic and unilateral cases of Peters’ anomaly could indicate that teratogenic, rather than genetic, factors are more important. However, it is interesting to note that eye defects in Foxe3 heterozygous mice have a penetrance of only 40% and are frequently unilateral. Incomplete penetrance and variable phenotype are characteristic of many haploinsufficiencies and are likely to reflect stochastic events at a bottleneck stage of embryonic development when the concentration of the gene product is critical. In the human population, interactions between different loci are likely to influence penetrance, but we do not think that this contributes to the variability we see in Foxe3 −/+ mice. Eye defects are equally common in offspring of affected and phenotypically normal Foxe3 heterozygotes. 
Establishment of a definite link between Peters’ anomaly and mutations in FOXE3 awaits analysis of more patients, segregation analysis in familial cases, and identification of obvious null alleles. However, the striking similarity between the phenotype of mouse Foxe3 heterozygotes and Peters’ anomaly in humans, combined with identification of a rare, nonconservative missense mutation affecting a conserved amino acid in the forkhead domain, suggests a causal relationship. 
 
Figure 1.
 
Human FOXE3. (a) Full-length human FOXE3 cDNA sequence compiled from four overlapping clones isolated from lens epithelium cDNA libraries. The amino acid sequence is predicted from the longest open reading frame, and the forkhead DNA binding domain is shaded. (b) Comparison of human and mouse FOXE3 amino acid sequences. Black background: identical amino acids; gray shading: conservative substitutions. The predicted human protein is 31 amino acids longer, because of a six-residue extension in the N terminus and expansion in four internal regions of low sequence complexity. Amino acids equivalent to the positions of four α-helices (H1–H4) and the two strands of a β-sheet (S2, S3) in FOXC2 (FREAC11 45 ) are indicated by boxes under the alignment. (☆) The two residues that are altered in the dyl mouse mutant; (★) residue substituted in patient LAS (see Fig. 5 ).
Figure 1.
 
Human FOXE3. (a) Full-length human FOXE3 cDNA sequence compiled from four overlapping clones isolated from lens epithelium cDNA libraries. The amino acid sequence is predicted from the longest open reading frame, and the forkhead DNA binding domain is shaded. (b) Comparison of human and mouse FOXE3 amino acid sequences. Black background: identical amino acids; gray shading: conservative substitutions. The predicted human protein is 31 amino acids longer, because of a six-residue extension in the N terminus and expansion in four internal regions of low sequence complexity. Amino acids equivalent to the positions of four α-helices (H1–H4) and the two strands of a β-sheet (S2, S3) in FOXC2 (FREAC11 45 ) are indicated by boxes under the alignment. (☆) The two residues that are altered in the dyl mouse mutant; (★) residue substituted in patient LAS (see Fig. 5 ).
Figure 2.
 
Foxe3 dyl encodes a protein that is defective in DNA binding. Gel-shift assay with a 32P-labeled forkhead consensus-binding site and in vitro-translated protein. FOXC1 (formerly FREAC3 or FKHL7) is a related forkhead protein that has previously been shown to bind this site with high affinity. Normal Foxe3 bound with an affinity similar to that of FOXC1, but no shifted complex was seen, with the mutant protein containing the two dyl substitutions. The negative control consisted of reticulocyte lysate mock translated without RNA.
Figure 2.
 
Foxe3 dyl encodes a protein that is defective in DNA binding. Gel-shift assay with a 32P-labeled forkhead consensus-binding site and in vitro-translated protein. FOXC1 (formerly FREAC3 or FKHL7) is a related forkhead protein that has previously been shown to bind this site with high affinity. Normal Foxe3 bound with an affinity similar to that of FOXC1, but no shifted complex was seen, with the mutant protein containing the two dyl substitutions. The negative control consisted of reticulocyte lysate mock translated without RNA.
Figure 3.
 
Slit lamp photographs of eyes of mice and a human. (a) Normal (wild-type) eye of an adult BALB/c mouse. (b) Eye of an adult Foxe3 dyl/+ heterozygous mouse. The central corneal opacity and unevenness were caused by a persistent connection between lens and cornea. (c) Eye of a patient with Peters’ anomaly.
Figure 3.
 
Slit lamp photographs of eyes of mice and a human. (a) Normal (wild-type) eye of an adult BALB/c mouse. (b) Eye of an adult Foxe3 dyl/+ heterozygous mouse. The central corneal opacity and unevenness were caused by a persistent connection between lens and cornea. (c) Eye of a patient with Peters’ anomaly.
Figure 4.
 
Histologic sections of adult (9-week-old) wild-type (a, d) and Foxe3 dyl/+ heterozygous (b, c, el) mice. (ac) Low-magnification view of a wild-type eye (a) and two mutant eyes (b, c), showing the two typical defects associated with Foxe3 heterozygosity: a connection between lens and cornea (b) and corneal swelling with cavities in the stroma (c; also seen on either side of the connection in b). (dl) Higher magnification views of wild-type (d) and heterozygous (el) eyes, showing the range of phenotypes observed in affected eyes. Foxe3 haploinsufficiency manifests itself as: adherence between lens and cornea, which can be direct (l) or through a stalk-like connection (e, f, j); a disturbed curvature of the cornea (e, f, hj, l); the presence of foreign material in the corneal stroma (f, i, j, l) or under the lens capsule (g, l); and a lax and swollen corneal structure with cavities in the stroma (e, h, k).
Figure 4.
 
Histologic sections of adult (9-week-old) wild-type (a, d) and Foxe3 dyl/+ heterozygous (b, c, el) mice. (ac) Low-magnification view of a wild-type eye (a) and two mutant eyes (b, c), showing the two typical defects associated with Foxe3 heterozygosity: a connection between lens and cornea (b) and corneal swelling with cavities in the stroma (c; also seen on either side of the connection in b). (dl) Higher magnification views of wild-type (d) and heterozygous (el) eyes, showing the range of phenotypes observed in affected eyes. Foxe3 haploinsufficiency manifests itself as: adherence between lens and cornea, which can be direct (l) or through a stalk-like connection (e, f, j); a disturbed curvature of the cornea (e, f, hj, l); the presence of foreign material in the corneal stroma (f, i, j, l) or under the lens capsule (g, l); and a lax and swollen corneal structure with cavities in the stroma (e, h, k).
Figure 5.
 
A mutation in FOXE3 in a patient with Peters’ anomaly. (a) Left eye of patient LAS showing central leucoma. (b) Sequence trace of a normal human FOXE3 allele. For positions in the predicted amino acid sequence, see Figure 1b . (c) Direct sequencing of a FOXE3 PCR product from patient LAS with Peters’ anomaly, showing a T/G heterozygosity, which indicates the presence of a G→T mutant allele that substitutes leucine for Arg90 in the forkhead domain. (d) Sequence trace of a plasmid clone containing the mutant allele of patient LAS.
Figure 5.
 
A mutation in FOXE3 in a patient with Peters’ anomaly. (a) Left eye of patient LAS showing central leucoma. (b) Sequence trace of a normal human FOXE3 allele. For positions in the predicted amino acid sequence, see Figure 1b . (c) Direct sequencing of a FOXE3 PCR product from patient LAS with Peters’ anomaly, showing a T/G heterozygosity, which indicates the presence of a G→T mutant allele that substitutes leucine for Arg90 in the forkhead domain. (d) Sequence trace of a plasmid clone containing the mutant allele of patient LAS.
The authors thank Shomi Bhattacharya and Ordan Lehmann for control DNA samples, Veronica van Heyningen for helpful advice, and Mariavittoria Sparacio for technical assistance. 
Jimenez-Sanchez G, Childs B, Valle D. Human disease genes. Nature. 2001;409:853–855. [CrossRef] [PubMed]
Kume T, Deng KY, Winfrey V, Gould DB, Walter MA, Hogan BL. The forkhead/winged helix gene Mf1 is disrupted in the pleiotropic mouse mutation congenital hydrocephalus. Cell. 1998;93:985–996. [CrossRef] [PubMed]
Nishimura DY, Swiderski RE, Alward WLM, et al. The forkhead transcription factor gene FKHL7 is responsible for glaucoma phenotypes which map to 6p25. Nat Genet. 1998;19:140–147. [CrossRef] [PubMed]
Mears AJ, Jordan T, Mirzayans F, et al. Mutations of the forkhead/winged-helix gene, FREAC3/FKHL7, in patients with Axenfeld-Rieger anomaly. Am J Hum Genet. 1998;63:1316–1328. [CrossRef] [PubMed]
Kidson SH, Kume T, Deng K, Winfrey V, Hogan BL. The forkhead/winged-helix gene, Mf1, is necessary for the normal development of the cornea and formation of the anterior chamber in the mouse eye. Dev Biol. 1999;211:306–322. [CrossRef] [PubMed]
Smith RS, Zabaleta A, Kume T, et al. Haploinsufficiency of the transcription factors FOXC1 and FOXC2 results in aberrant ocular development. Hum Mol Genet. 2000;9:1021–1032. [CrossRef] [PubMed]
Winnier GE, Hargett L, Hogan BL. The winged helix transcription factor MFH1 is required for proliferation and patterning of paraxial mesoderm in the mouse embryo. Genes Dev. 1997;11:926–940. [CrossRef] [PubMed]
Iida K, Koseki H, Kakinuma H, et al. Essential roles of the winged helix transcription factor MFH-1 in aortic arch patterning and skeletogenesis. Development. 1997;124:4627–4638. [PubMed]
Fang J, Dagenais SL, Erickson RP, et al. Mutations in FOXC2 (MFH-1), a forkhead family transcription factor, are responsible for the hereditary lymphedema-distichiasis syndrome. Am J Hum Genet. 2000;67:1382–1388. [CrossRef] [PubMed]
Semina EV, Reiter R, Leysens NJ, et al. Cloning and characterization of a novel bicoid-related homeobox transcription factor gene, RIEG, involved in Rieger syndrome. Nat Genet. 1996;14:392–399. [CrossRef] [PubMed]
Ton CC, Hirvonen H, Miwa H, et al. Positional cloning and characterization of a paired box- and homeobox- containing gene from the aniridia region. Cell. 1991;67:1059–1074. [CrossRef] [PubMed]
Hanson IM, Fletcher JM, Jordan T, et al. Mutations at the PAX6 locus are found in heterogeneous anterior segment malformations including Peters” anomaly. Nat Genet. 1994;6:168–173. [CrossRef] [PubMed]
Hill RE, Favor J, Hogan BL, et al. Mouse small eye results from mutations in a paired-like homeobox-containing gene. [published correction appears in Nature 1992;355:750]Nature. 1991;354:522–525. [CrossRef] [PubMed]
Glaser T, Jepeal L, Edwards JG, Young SR, Favor J, Maas RL. PAX6 gene dosage effect in a family with congenital cataracts, aniridia, anophthalmia and central nervous system defects. Nat Genet. 1994;7:463–471. [CrossRef] [PubMed]
Jordan T, Hanson I, Zaletayev D, et al. The human PAX6 gene is mutated in two patients with aniridia. Nat Genet. 1992;1:328–332. [CrossRef] [PubMed]
Sisodiya SM, Free SL, Williamson KA, et al. PAX6 haploinsufficiency causes cerebral malformation and olfactory dysfunction in humans. Nat Genet. 2001;28:214–216. [CrossRef] [PubMed]
Hogan BL, Horsburgh G, Cohen J, Hetherington CM, Fisher G, Lyon MF. Small eyes (Sey): a homozygous lethal mutation on chromosome 2 which affects the differentiation of both lens and nasal placodes in the mouse. J Embryol Exp Morphol. 1986;97:95–110. [PubMed]
Theiler K, Varnum DS, Stevens LC. Development of Dickie’s small eye, a mutation in the house mouse. Anat Embryol (Berl). 1978;155:81–86. [PubMed]
Grindley JC, Davidson DR, Hill R. E. The role of Pax-6 in eye and nasal development. Development. 1995;121:1433–1442. [PubMed]
Lehmann OJ, Ebenezer ND, Jordan T, et al. Chromosomal duplication involving the forkhead transcription factor gene FOXC1 causes iris hypoplasia and glaucoma. Am J Hum Genet. 2000;67:1129–1135. [CrossRef] [PubMed]
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. [CrossRef] [PubMed]
Larsson C, Hellqvist M, Pierrou S, White I, Enerbäck S, Carlsson P. Chromosomal localization of six human forkhead genes, FREAC-1 (FKHL5), -3 (FKHL7), -4 (FKHL8), -5 (FKHL9), -6 (FKHL10), and -8 (FKHL12). Genomics. 1995;30:464–469. [CrossRef] [PubMed]
Blixt Å, Mahlapuu M, Aitola M, Pelto-Huikko M, Enerbäck S, Carlsson P. A forkhead gene, FoxE3, is essential for lens epithelial proliferation and closure of the lens vesicle. Genes Dev. 2000;14:245–254. [PubMed]
Brownell I, Dirksen M, Jamrich M. Forkhead Foxe3 maps to the dysgenetic lens locus and is critical in lens development and differentiation. Genesis. 2000;27:81–93. [CrossRef] [PubMed]
Sanyal S, Hawkins RK. Dysgenetic lens (dyl): a new gene in the mouse. Invest Ophthalmol Vis Sci. 1979;18:642–645. [PubMed]
Semina EV, Brownell I, Mintz-Hittner HA, Murray JC, Jamrich M. Mutations in the human forkhead transcription factor FOXE3 associated with anterior segment ocular dysgenesis and cataracts. Hum Mol Genet. 2001;10:231–236. [CrossRef] [PubMed]
Doward W, Perveen R, Lloyd IC, Ridgway AE, Wilson L, Black GC. A mutation in the RIEG1 gene associated with Peters” anomaly. J Med Genet. 1999;36:152–155. [PubMed]
Vincent A, Billingsley G, Priston M, et al. Phenotypic heterogeneity of CYP1B1: mutations in a patient with Peters” anomaly. J Med Genet. 2001;38:324–326. [CrossRef] [PubMed]
Churchill AJ, Booth AP, Anwar R, Markham AF. PAX 6 is normal in most cases of Peters” anomaly. Eye. 1998;12:299–303. [CrossRef] [PubMed]
Pierrou S, Hellqvist M, Samuelsson L, Enerbäck S, Carlsson P. Cloning and characterization of seven human forkhead proteins: binding site specificity and DNA bending. EMBO J. 1994;13:5002–5012. [PubMed]
Lander E, Linton S, Birren LM, et al. Initial sequencing and analysis of the human genome. Nature. 2001;409:860–921. [CrossRef] [PubMed]
Venter J, Adams C, Myers MD, et al. The sequence of the human genome. Science. 2001;291:1304–1351. [CrossRef] [PubMed]
Pierrou S, Enerbäck S, Carlsson P. Selection of high-affinity binding sites for sequence-specific, DNA binding proteins from random sequence oligonucleotides. Anal Biochem. 1995;229:99–105. [CrossRef] [PubMed]
Peters’ A. Über angeboren Defektbildung der Descementschen Membran. Klin Mbl Augenheilk. 1906;44:27–40.
Stone DL, Kenyon KR, Green WR, Ryan SJ. Congenital central corneal leukoma (Peters” anomaly). Am J Ophthalmol. 1976;81:173–193. [CrossRef] [PubMed]
Sarfarazi M, Stoilov I. Molecular genetics of primary congenital glaucoma. Eye. 2000;14:422–428. [CrossRef] [PubMed]
Saleem RA, Banerjee-Basu S, Berry FB, Baxevanis AD, Walter MA. Analyses of the effects that disease-causing missense mutations have on the structure and function of the winged-helix protein FOXC1. Am J Hum Genet. 2001;68:627–641. [CrossRef] [PubMed]
Easty DL, Sparrow JM. Oxford Textbook of Ophthalmology. 1999; Oxford University Press New York.
Kirkness CM, McCartney A, Rice NS, Garner A, Steele AD. Congenital hereditary corneal oedema of Maumenee: its clinical features, management, and pathology. Br J Ophthalmol. 1987;71:130–144. [CrossRef] [PubMed]
Mullaney PB, Risco JM, Teichmann K, Millar L. Congenital hereditary endothelial dystrophy associated with glaucoma. Ophthalmology. 1995;102:186–192. [CrossRef] [PubMed]
Coulombre AJ, Coulombre JL. Lens development: the role of the lens in eye growth. J Exp Zool. 1964;156:39–48. [CrossRef] [PubMed]
Genis-Galvez JM. Role of the lens in the morphogenesis of the iris and cornea. Nature. 1966;210:209–210. [CrossRef] [PubMed]
Genis-Galvez JM, Santos-Gutierrez L, Rios-Gonzalez A. Causal factors in corneal development: an experimental analysis in the chick embryo. Exp Eye Res. 1967;6:48–56. [CrossRef] [PubMed]
Clark KL, Halay ED, Lai E, Burley SK. Co-crystal structure of the HNF-3/fork head DNA-recognition motif resembles histone H5. Nature. 1993;364:412–420. [CrossRef] [PubMed]
van Dongen MJ, Cederberg A, Carlsson P, Enerback S, Wikstrom M. Solution structure and dynamics of the DNA-binding domain of the adipocyte-transcription factor FREAC-11. J Mol Biol. 2000;296:351–359. [CrossRef] [PubMed]
Hellqvist M, Mahlapuu M, Blixt Å, Enerbäck S, Carlsson P. The human forkhead protein FREAC-2 contains two functionally redundant activation domains and interacts with TBP and TFIIB. J Biol Chem. 1998;273:23335–23343. [CrossRef] [PubMed]
Qian X, Costa RH. Analysis of hepatocyte nuclear factor-3 beta protein domains required for transcriptional activation and nuclear targeting. Nucleic Acids Res. 1995;23:1184–1191. [CrossRef] [PubMed]
Figure 1.
 
Human FOXE3. (a) Full-length human FOXE3 cDNA sequence compiled from four overlapping clones isolated from lens epithelium cDNA libraries. The amino acid sequence is predicted from the longest open reading frame, and the forkhead DNA binding domain is shaded. (b) Comparison of human and mouse FOXE3 amino acid sequences. Black background: identical amino acids; gray shading: conservative substitutions. The predicted human protein is 31 amino acids longer, because of a six-residue extension in the N terminus and expansion in four internal regions of low sequence complexity. Amino acids equivalent to the positions of four α-helices (H1–H4) and the two strands of a β-sheet (S2, S3) in FOXC2 (FREAC11 45 ) are indicated by boxes under the alignment. (☆) The two residues that are altered in the dyl mouse mutant; (★) residue substituted in patient LAS (see Fig. 5 ).
Figure 1.
 
Human FOXE3. (a) Full-length human FOXE3 cDNA sequence compiled from four overlapping clones isolated from lens epithelium cDNA libraries. The amino acid sequence is predicted from the longest open reading frame, and the forkhead DNA binding domain is shaded. (b) Comparison of human and mouse FOXE3 amino acid sequences. Black background: identical amino acids; gray shading: conservative substitutions. The predicted human protein is 31 amino acids longer, because of a six-residue extension in the N terminus and expansion in four internal regions of low sequence complexity. Amino acids equivalent to the positions of four α-helices (H1–H4) and the two strands of a β-sheet (S2, S3) in FOXC2 (FREAC11 45 ) are indicated by boxes under the alignment. (☆) The two residues that are altered in the dyl mouse mutant; (★) residue substituted in patient LAS (see Fig. 5 ).
Figure 2.
 
Foxe3 dyl encodes a protein that is defective in DNA binding. Gel-shift assay with a 32P-labeled forkhead consensus-binding site and in vitro-translated protein. FOXC1 (formerly FREAC3 or FKHL7) is a related forkhead protein that has previously been shown to bind this site with high affinity. Normal Foxe3 bound with an affinity similar to that of FOXC1, but no shifted complex was seen, with the mutant protein containing the two dyl substitutions. The negative control consisted of reticulocyte lysate mock translated without RNA.
Figure 2.
 
Foxe3 dyl encodes a protein that is defective in DNA binding. Gel-shift assay with a 32P-labeled forkhead consensus-binding site and in vitro-translated protein. FOXC1 (formerly FREAC3 or FKHL7) is a related forkhead protein that has previously been shown to bind this site with high affinity. Normal Foxe3 bound with an affinity similar to that of FOXC1, but no shifted complex was seen, with the mutant protein containing the two dyl substitutions. The negative control consisted of reticulocyte lysate mock translated without RNA.
Figure 3.
 
Slit lamp photographs of eyes of mice and a human. (a) Normal (wild-type) eye of an adult BALB/c mouse. (b) Eye of an adult Foxe3 dyl/+ heterozygous mouse. The central corneal opacity and unevenness were caused by a persistent connection between lens and cornea. (c) Eye of a patient with Peters’ anomaly.
Figure 3.
 
Slit lamp photographs of eyes of mice and a human. (a) Normal (wild-type) eye of an adult BALB/c mouse. (b) Eye of an adult Foxe3 dyl/+ heterozygous mouse. The central corneal opacity and unevenness were caused by a persistent connection between lens and cornea. (c) Eye of a patient with Peters’ anomaly.
Figure 4.
 
Histologic sections of adult (9-week-old) wild-type (a, d) and Foxe3 dyl/+ heterozygous (b, c, el) mice. (ac) Low-magnification view of a wild-type eye (a) and two mutant eyes (b, c), showing the two typical defects associated with Foxe3 heterozygosity: a connection between lens and cornea (b) and corneal swelling with cavities in the stroma (c; also seen on either side of the connection in b). (dl) Higher magnification views of wild-type (d) and heterozygous (el) eyes, showing the range of phenotypes observed in affected eyes. Foxe3 haploinsufficiency manifests itself as: adherence between lens and cornea, which can be direct (l) or through a stalk-like connection (e, f, j); a disturbed curvature of the cornea (e, f, hj, l); the presence of foreign material in the corneal stroma (f, i, j, l) or under the lens capsule (g, l); and a lax and swollen corneal structure with cavities in the stroma (e, h, k).
Figure 4.
 
Histologic sections of adult (9-week-old) wild-type (a, d) and Foxe3 dyl/+ heterozygous (b, c, el) mice. (ac) Low-magnification view of a wild-type eye (a) and two mutant eyes (b, c), showing the two typical defects associated with Foxe3 heterozygosity: a connection between lens and cornea (b) and corneal swelling with cavities in the stroma (c; also seen on either side of the connection in b). (dl) Higher magnification views of wild-type (d) and heterozygous (el) eyes, showing the range of phenotypes observed in affected eyes. Foxe3 haploinsufficiency manifests itself as: adherence between lens and cornea, which can be direct (l) or through a stalk-like connection (e, f, j); a disturbed curvature of the cornea (e, f, hj, l); the presence of foreign material in the corneal stroma (f, i, j, l) or under the lens capsule (g, l); and a lax and swollen corneal structure with cavities in the stroma (e, h, k).
Figure 5.
 
A mutation in FOXE3 in a patient with Peters’ anomaly. (a) Left eye of patient LAS showing central leucoma. (b) Sequence trace of a normal human FOXE3 allele. For positions in the predicted amino acid sequence, see Figure 1b . (c) Direct sequencing of a FOXE3 PCR product from patient LAS with Peters’ anomaly, showing a T/G heterozygosity, which indicates the presence of a G→T mutant allele that substitutes leucine for Arg90 in the forkhead domain. (d) Sequence trace of a plasmid clone containing the mutant allele of patient LAS.
Figure 5.
 
A mutation in FOXE3 in a patient with Peters’ anomaly. (a) Left eye of patient LAS showing central leucoma. (b) Sequence trace of a normal human FOXE3 allele. For positions in the predicted amino acid sequence, see Figure 1b . (c) Direct sequencing of a FOXE3 PCR product from patient LAS with Peters’ anomaly, showing a T/G heterozygosity, which indicates the presence of a G→T mutant allele that substitutes leucine for Arg90 in the forkhead domain. (d) Sequence trace of a plasmid clone containing the mutant allele of patient LAS.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×