Molecular Characterization of Pax6^2Neu Through Pax6^10Neu: An Extension of the Pax6 Allelic Series and the Identification of Two Possible Hypomorph Alleles in the Mouse Mus musculus

Corresponding author: Jack Favor, Institute of Mammalian Genetics, GSF-Research Center for Environment and Health, Ingolstädter Landstr. 1, D-85764, Neuherberg, Germany., favor{at}gsf.de (E-mail)

Communicating editor: C. KOZAK

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Phenotype-based mutagenesis experiments will increase the mouse mutant resource, generating mutations at previously unmarked loci as well as extending the allelic series at known loci. Mapping, molecular characterization, and phenotypic analysis of nine independent Pax6 mutations of the mouse recovered in mutagenesis experiments is presented. Seven mutations result in premature termination of translation and all express phenotypes characteristic of null alleles, suggesting that Pax6 function requires all domains to be intact. Of major interest is the identification of two possible hypomorph mutations: Heterozygotes express less severe phenotypes and homozygotes develop rudimentary eyes and nasal processes and survive up to 36 hr after birth. Pax6^4Neu results in an amino acid substitution within the third helix of the homeodomain. Three-dimensional modeling indicates that the amino acid substitution interrupts the homeodomain recognition -helix, which is critical for DNA binding. Whereas cooperative dimer binding of the mutant homeodomain to a paired-class DNA target sequence was eliminated, weak monomer binding was observed. Thus, a residual function of the mutated homeodomain may explain the hypomorphic nature of the Pax6^4Neu allele. Pax6^7Neu is a base pair substitution in the Kozak sequence and results in a reduced level of Pax6 translation product. The Pax6^4Neu and Pax6^7Neu alleles may be very useful for gene-dosage studies.

IN the mouse a number of large-scale mutagenesis screens have been undertaken to systematically recover mutations with relevant phenotypes to complement the forthcoming genome sequence data and to provide the genetic variants to initiate functional genetic studies (HRABE DE ANGELIS et al. 2000 ; NOLAN et al. 2000 ). We have conducted an extensive series of mutagenesis experiments to recover dominant mutations that affect eye morphology (FAVOR and NEUHAUSER-KLAUS 2000 ). A total of 192 independent mutations are available for genetic and molecular analyses, which represents one of the largest collections of mutations affecting the development and function of a defined organ. We are systematically mapping all recovered mutations to identify the genes/loci responsible for eye development. Our results to date indicate that the most mutable locus is at chromosome 2, centimorgan 58 in the vicinity of the Pax6 gene, which is known to function in eye development. This combination of chromosomal location and the known involvement of Pax6 in eye development suggested that Pax6 is the candidate gene affected in this large group of mutations.

The mouse Pax6 gene belongs to the family of paired-box-containing genes, which function as transcription factors regulating developmental processes. Pax6 encodes a protein with both a paired domain and homeodomain, separated by a linker segment, and followed by a C-terminal proline-, serine-, and threonine-rich region (WALTHER and GRUSS 1991 ). The highly conserved paired domain and homeodomain are DNA-binding domains while the proline-, serine-, and threonine-rich domain functions in transcriptional activation (GLASER et al. 1994 ). Mutant analyses have shown that Pax6 plays a role in the development of the eye (THEILER et al. 1978 ; HOGAN et al. 1988 ), nasal derivatives (HOGAN et al. 1988 ; HEINZMANN et al. 1991 ; GRINDLEY et al. 1995 ; QUINN et al. 1996 ), additional craniofacial traits (KAUFMAN et al. 1995 ), the central nervous system (CNS; SCHMAHL et al. 1993 ; GRINDLEY et al. 1997 ; STOYKOVA et al.. 1996 , STOYKOVA et al.. 1997 , STOYKOVA et al.. 2000 ; GOTZ et al. 1998 ), the pancreas (ST-ONGE et al. 1997 ), and the pituitary gland (BENTLEY et al. 1999 ; KIOUSSI et al. 1999 ). Heterozygous carriers of Pax6 null mutations express phenotypic effects, which suggests that a defined concentration of Pax6 protein activity is necessary for normal development (HILL et al. 1991 ). This suggestion is supported by studies in which the wild-type Pax6 allele was overexpressed and resulted in abnormal eye development (SCHEDL et al. 1996 ).

Human patients heterozygous for PAX6 mutant alleles express eye abnormalities similar to those observed in the mouse (BROWN et al. 1998 ; PROSSER and VAN HEYNINGEN 1998 ). Most described PAX6 mutant alleles result in premature termination of translation. The underrepresentation of missense mutations may be due to the fact that hypomorph alleles likely result in a less severe phenotype and patients who carry hypomorph alleles might not have been included in a mutation screen.

The identification of Pax6 hypomorph alleles in a laboratory animal would be useful to experimentally address the question of Pax6 protein activity dosage and phenotypic defects. Here we report nine new Pax6 mutant alleles in the mouse. Seven mutations result in a truncated gene product. Of major interest is the identification of two possible hypomorph alleles. One is an amino acid substitution in the homeodomain. The second is a base pair substitution in the Kozak sequence.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Animals:
All inbred strain (C3H, C57BL/6, 102), F₁ hybrid (102xC3H) and Tester-stock animals were obtained from breeding colonies maintained by the GSF-Department of Animal Resources at Neuherberg.

Mutation recovery:
Mutants were recovered in the offspring of (102xC3H)F₁ hybrid or C3H males that were mutagenically treated and mated to untreated Tester-stock or C3H females, respectively. Prior to the initiation of the mutagenesis experiments, all parental animals were ophthalmologically examined and only those that did not express an eye morphological abnormality were employed in order to eliminate any preexisting eye mutations.

Ophthalmological examination:
Mice were examined biomicroscopically for eye abnormalities at weaning. Pupils were dilated with a 1% atropin solution applied to the eyes at least 10 min prior to examination. Both eyes of the mice were examined by slit lamp biomicroscopy (Zeiss SLM30) at x48 magnification with a narrow beam slit lamp illumination at a 25°–30° angle from the direction of observation.

Genetic confirmation crosses:
The recovered presumed dominant eye mutants were outcrossed to wild-type strain 102, C3H, or (102xC3H)F₁ partners and at least 20 offspring were ophthalmologically examined for transmission of the specific eye abnormality. The offspring were carefully examined to arrive at a more complete description of the phenotype expressed by heterozygous carriers. A detailed consideration of the ophthalmological examination procedures and genetic confirmation test has been published (FAVOR 1983 ). All confirmed mutations were maintained by outcrossing heterozygous carriers to homozygous wild-type partners. Before initiating the mapping experiments each mutation was backcrossed at least 10 generations to strain C3H.

Mapping and allelism tests:
Complete details of our mapping procedures have been published recently (FAVOR et al. 1997 ). Segregation data were analyzed with Map Manager, Version 2.6.5 (MANLY 1993 ), and the gene order was determined by minimizing the number of multiple crossovers. Allelism was tested by mating heterozygous carriers of an unmapped mutation (ENU-65 or ENU-5011) to heterozygous Pax6^Sey-Neu/+ mice. Pregnant females were dissected at stages E15 to E17 and embryos were classified for the typical homozygote mutant Pax6 phenotype indicative of noncomplementation (anophthalmia and craniofacial abnormalities).

RT-PCR and sequence analysis:
Total RNA was isolated from heads of E15 embryos expressing the typical wild-type or homozygous mutant Pax6 phenotype recovered in inter se crosses of heterozygotes using the RNeasy kit (QIAGEN, Chatsworth, CA). RNA was reverse transcribed and the entire coding region was PCR amplified with a Titan one Tube RT-PCR system (Roche Diagnostics, Mannheim, Germany) as recommended by the supplier. Two overlapping primer sets, A, 5'-CAGAAGACTTTAACCAAGGGC [nucleotide (nt) 33–53] and 5'-ATCCTTAGTTTATCATACATGCCG (nt 647–623), and B, 5'-AACAGAGTTCTTCGCAACCTGG (nt 573–595) and 5'-GCTGTGTCCACATAGTCATTGGC (nt 1545–1522), were used for the amplification of the Pax6 cDNA. The PCR products were HPLC purified and sequenced with a Taq Dye-Deoxy terminator cycle sequencing kit by SequiServe (Vaterstetten, Germany). Additional primers were used for the complete sequencing of the coding region as well as for the confirmation of the mutation employing genomic DNA as a template.

Numbering of the Pax6 nucleotides and codons/amino acids follows that of WALTHER and GRUSS 1991 .

Assay for nasal development:
To assay the extent of nasal development in Pax6 mutant mice we utilized the Pax9^lacZ reporter gene, which is expressed in nasal mesenchyme (PETERS et al. 1998 ). Double heterozygotes, consisting of Pax9^lacZ/+ with Pax6^Sey-Neu/+ or Pax6^4Neu/+, were constructed and bred to heterozygous carriers of the respective Pax6 mutant allele. Embryos (E12) were identified as homozygous wild type and heterozygous or homozygous mutant at the Pax6 locus by phenotype. Heterozygous Pax9^lacZ carriers were identified by a PCR assay as previously described (PETERS et al. 1998 ). Whole mount Pax9^lacZ/+ embryos were stained with X-Gal for 8–12 hr according to established protocols (GOSSLER and ZACHGO 1993 ). To completely visualize Pax9^lacZ expression, embryos were dehydrated in methanol and cleared with benzyl benzoate/benzyl alcohol.

Assay for eye phenotype:
Offspring from the cross-mutant heterozygote with inbred strain C3H/J were ophthalmologically examined at 21 days of age and categorized for the degree of lens/corneal opacity. At 30 days of age the offspring were sacrificed by cervical dislocation, and eyes were enucleated, washed once in PBS, rinsed in H₂O, blotted dry on filter paper, and weighed. Eye weight data were statistically analyzed by Factorial ANOVA, General Linear Models Procedure, employing SAS Software release 6.12 (Cary, NC).

Histology:
Pregnant females from Pax6^4Neu/+ or Pax6^7Neu/+ inter se crosses were sacrificed on day 16 postconception and embryos assigned a genotype on the basis of phenotype (heterozygotes have a slightly smaller eye with distorted shape of the pupil; homozygous mutants are anophthalmic). Embryos were fixed in 10% buffered formalin, and heads were embedded in paraffin and serially sectioned (coronal) at 5 µm. Sections were stained with hematoxylin and eosin.

Three-dimensional modeling:
A three-dimensional model of the Pax6 homeodomain (Leu1 through Glu60) was constructed by homology modeling using the atomic coordinates of the homeodomains of the genes paired (WILSON et al. 1995 ), Antennapedia (QIAN et al. 1994B ), VND/NK-2 (TSAO et al. 1995 ), engrailed (KISSINGER et al. 1990 ), and Fushi Tarazu (QIAN et al. 1994A ). Protein sequences were aligned with FASTA (PEARSON 1990 ). Three-dimensional modeling was performed with the InsightII software (Molecular Simulations, San Diego) following a previously described procedure (KRUGER et al. 1998 ). The model of the Pax6 homeodomain was docked to its DNA recognition site, guided by the atomic coordinates of the paired/DNA complex (WILSON et al. 1995 ).

Expression of Pax6 homeodomains of the wild-type and Pax6^4Neu alleles in bacteria:
The appropriate fragments were amplified from cDNA templates and inserted between the PstI and HindIII sites of the pQE-41 vector (QIAGEN). The expression constructs result in the in-frame fusion of the desired sequence (the Pax6 amino acids 218 to 288, which includes the homeodomain and six amino acid residues upstream and four amino acid residues downstream) to the C-terminal end of the mouse dihydrofolate reductase protein. The pQE expression constructs were transformed into Escherichia coli strain M15 (pREP4), which carries the lacI^q repressor gene. Cultures of exponentially growing bacteria were induced with 0.1 mM isopropyl thiogalactoside for 2 hr at 30°. Bacterial pellets were lysed in 50 mM NaH₂PO₄, 300 mM NaCl, pH 8.0, 0.5 mg/ml lysozyme, 5 µg/ml DNaseI and 1 mM phenylmethylsulfonyl fluoride. Crude extracts were electrophoresed in 10% SDS-PAGE and proteins were visualized by staining with Coomassie brilliant blue.

Western blotting:
Crude extracts were fractionated on a 10% SDS-PAGE and then electrophoretically transferred to fluorotrans membranes (Pall Europe Ltd, Portsmouth, UK) and probed with rabbit anti-Pax6-homeodomain antiserum (1:200; serum 13, gift from Dr. S. Saule, Institut Pasteur de Lille), which was detected with anti-rabbit IgG alkaline phosphatase conjugate (1:1000; Promega, Madison, WI).

Electrophoretic mobility shift analysis:
Crude extracts from the transformed bacteria were incubated with 15 fmol of the target oligonucleotide, which was 3' end-labeled with digoxygenin-11-ddUTP as recommended by the supplier (Roche Diagnostics, Mannheim, Germany). The target oligonucleotide contained the Pax6 homeodomain binding sequence P3 (CZERNY and BUSSLINGER 1995 ), in which the two halves of the palindromic sequence were separated by TGA as in the rhodopsin promoter (SHENG et al. 1997 ). The sequence of the double-strand oligonucleotide (with P3 underlined)is as follows: 5'-TCGAGGGCATCAGGATGCTAATTGAATTAGCATCCGATCGGG-3' and 3'-CCCGTAGTCCTACGATTAACTTAATCGTAGGCTAGCCCAGCT-5'.

Protein-DNA complexes were separated on a native 10% polyacrylamide gel (in 0.25x Tris-borate-EDTA), blotted onto BM nylon membranes, and detected by autoradiography following an enzyme immunoassay with antidigoxygenin (Fab)-alkaline phosphatase and the chemiluminiscent substrate CSPD (Roche).

The sequence data presented in this article have been submitted to the EMBL/GenBank Data Libraries under the accession numbers: Pax6^2Neu, Y19193; Pax6^3Neu, Y19195; Pax6^4Neu, Y19196; Pax6^5Neu, Y19197; Pax6^6Neu, Y19198; Pax6^7Neu, Y19199; Pax6^9Neu, AJ292077; and Pax6^10Neu, AJ307468.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Mutations:
Table 1 gives the original mutant identification, the proposed allele symbol (based on the molecular characterization results below), the mutagenic treatment of the parental male mice, and reference to the mutagenesis experiments in which the original mutants were recovered. Seven mutations (ENU-5011, ENU-5045, ENU-636, ENU-2033, ENU-642, TX-12, and ENU-6037) were recovered in offspring derived from fertilizations in which the participating parental male germ cells were mutagenically treated at the stem cell spermatogonial stage (fertilizations occurred >49 days after treatment). The original mutant recovered for each of these independent mutations expressed bilateral anterior polar cataract, corneal adhesions, and, depending on the mutant allele, slight to more severe microphthalmia. Genetic confirmation tests indicated that the original mutants were heterozygous for an autosomal dominant allele. Two mutations (ENU-65 and nPMS-9) were recovered in offspring derived from fertilizations in which the participating parental male germ cells were mutagenically exposed at the spermatocyte stage. The fertilization producing the mutant ENU-65 occurred 24 days after treatment. The fertilization producing the mutant nPMS-9 occurred 20 days after treatment. The original mutant ENU-65 expressed unilaterally anterior polar opacity, corneal adhesions, and microphthalmia. Genetic confirmation and subsequent breeding results suggested that the original mutant was a gonosomic mosaic for an autosomal dominant mutation (FAVOR 1983 ; FAVOR et al. 1990 ; FAVOR and NEUHAUSER-KLAUS 1994 ). The original mutant nPMS-9 expressed bilateral eye defects and genetic confirmation crosses were consistent with the assumption that the mutant was heterozygous for an autosomal dominant mutation.

The expressivity of all mutations was variable with mutant carriers expressing a range of phenotypes from small anterior polar cataracts to the more extreme phenotype of anterior polar opacity, corneal adhesions, iris abnormalities, and microphthalmia. Furthermore, the degree of phenotype expressed between the eyes of an individual mutation carrier was variable.

Examination of the eyes of animals not treated with atropin revealed that heterozygous carriers for all mutant alleles expressed greatly reduced or absent iris.

Breeding results are given in Table 2. Segregation analyses for the outcross of heterozygous carriers to wild type indicated a 1:1 ratio of wild type to mutant carrier offspring consistent with an autosomal dominant mutation without viability or penetrance effects for six mutations (Pax6^2Neu, Pax6^3Neu, Pax6^4Neu, Pax6^5Neu, Pax6^6Neu, and Pax6^10Neu). The segregation analyses of three mutations (Pax6^7Neu, Pax6^8Neu, and Pax6^9Neu) deviated significantly from the expected 1:1 ratio. For Pax6^8Neu and Pax6^9Neu there was an excess of wild-type offspring relative to mutants. The deviations in segregation observed for these two mutations may be due to a conservative classification of offspring in the initial generations of breeding or to genetic background effects since, as will be seen below, (a) Pax6^8Neu (with an observed deviation in segregation) is an exact repeat mutation of that carried by Pax6^6Neu (with no deviation in segregation) and (b) once the spectrum of phenotypes associated with the mutations became known we were able to classify the genotypes on the basis of phenotype with no deviation in the segregation ratio in the mapping studies. The segregation results for Pax6^7Neu showed an excess of mutant offspring. However, there was no deviation in the segregation ratio in the mapping analyses and we conclude the deviation to be spurious.

Examination of embryos from inter se crosses of heterozygous carriers of seven mutations (Pax6^2Neu, Pax6^3Neu, Pax6^5Neu, Pax6^6Neu, Pax6^8Neu, Pax6^9Neu, and Pax6^10Neu) revealed that the resultant phenotypes were identical to those previously described (HOGAN et al. 1988 ): Heterozygotes express obvious microphthalmia with a triangular-shaped pupil and homozygotes express anophthalmia with severe craniofacial defects (Fig 2A). Embryos from two mutations deviated from this general phenotypic classification. Pax6^4Neu and Pax6^7Neu heterozygotes expressed abnormal pupil shape but eye size reduction was only slight; homozygous mutants were anophthalmic but some were observed to have a distinct optic pit with underlying pigmentation (Fig 2B and Fig C).

View larger version (25K): In this window In a new window Download PPT slide   Figure 1. Predicted protein product of wild-type and mutant Pax6 alleles. PD, paired domain; L, linker region; HD, homeodomain; P/S/T, proline-/serine-/threonine-rich region (the codon positions at the domain borders of the Pax6 wild type are shaded numbers); shaded boxes, aberrant amino acid sequences; *, missense mutation; dashed lines, reduced translation of Pax67Neu messenger RNA.

View larger version (64K): In this window In a new window Download PPT slide   Figure 2. Phenotypic characterization of wild-type, heterozygous, and homozygous mutant Pax6 embryos. (A) Homozygous wild-type, heterozygous, and homozygous mutant Pax63Neu embryos (E15) expressing the typical phenotypes for a Pax6 null allele. Heterozygotes have a clearly smaller eye with a distinctive triangular shaped pupil. Homozygous mutants are anophthalmic. (B) Pax64Neu embryos (E15) expressing the phenotypes for a Pax6 hypomorph allele. Heterozygotes are distinguishable from wild types only by the slight distortion in the shape of the eye. Homozygous mutants develop a rudimentary, pigmented eye (arrow). (C) Assay for nasal development in homozygous Pax6 mutant embryos (E12) employing the expression of the Pax9lacZ reporter gene in nasal mesenchyme of Pax9lacZ/+ embryos. As compared to homozygous wild type, homozygotes for the hypomorph Pax64Neu allele have reduced lateral (ln) and medial (mn) nasal processes. The rudimentary, pigmented eye in the Pax64Neu homozygote (arrows) can also be seen. The lateral and medial nasal processes are completely absent in homozygotes for the Pax6Sey-Neu null allele (right). ey, eye.

Mapping:
Two mutations (Pax6^2Neu and Pax6^3Neu) were shown to be alleles of Pax6 on the basis of noncomplementation with Pax6^Sey-Neu. Phenotypic analyses of the mutation Pax6^10Neu suggested that Pax6 was mutated. The remaining six mutations were localized relative to the chromosome 2 markers D2Mit249 and Agouti (Table 3). All mutations mapped to the same region, expressed a similar phenotype, and, therefore, were assumed to be mutant alleles of the same gene. Combined mapping results yielded the following gene order: D2Mit249-(1.4 cM)-Pax6-(25.3 cM)-Agouti.

Molecular characterization:
Sequence analyses indicated that the Pax6 gene was affected in all nine mutations (Table 4). Seven mutations (Pax6^2Neu, Pax6^4Neu, Pax6^5Neu, Pax6^6Neu, Pax6^7Neu, Pax6^8Neu, and Pax6^10Neu) were base pair substitutions, one mutation (Pax6^3Neu) was a base pair insertion, and one mutation (Pax6^9Neu) was a 7-bp deletion.

The predicted Pax6 protein products are given in Fig 1. Seven mutations (Pax6^2Neu, Pax6^3Neu, Pax6^5Neu, Pax6^6Neu, Pax6^8Neu, Pax6^9Neu, and Pax6^10Neu) result in a truncated Pax6 protein product. Pax6^2Neu is a base pair substitution at nucleotide position 2 of the intron 9 donor splice site. The isolated cDNA was 87 bp longer than wild type due to the destruction of the splice site and the use of a cryptic splice site. Translation of the aberrant mRNA is predicted to result in a normal amino acid sequence up to codon 269 of exon 9 followed by 23 amino acids encoded by the included intron 9 sequence and a premature stop. The mutation results in the truncation of the last 15 amino acids of the C terminus of the homeodomain and the entire P/S/T domain. The Pax6^3Neu frameshift mutation results in a normal amino acid sequence through codon 145 of exon 7 followed by 12 aberrant amino acids and a premature stop. The gene product contains a normal paired domain but the linker region, the homeodomain, and the P/S/T domain are all deleted. The base pair substitution in the Pax6^5Neu mutation creates a premature stop site in the C terminus of the paired box, resulting in a predicted gene product that is truncated for the last 27 amino acids of the paired domain as well as the linker region, the homeodomain, and the P/S/T domain. Pax6^6Neu is also a base pair substitution, which creates a premature stop codon and results in a truncated gene product from the N-terminal end of the P/S/T domain. Pax6^8Neu is a repeat mutation such as that observed in Pax6^6Neu. The 7-bp deletion associated with Pax6^9Neu is predicted to result in an extremely truncated gene product consisting of the first 32 amino acids followed by 17 aberrant amino acids and a premature stop. Finally, the Pax6^10Neu base pair substitution creates a stop codon in the C terminus of the paired box. The predicted mutant gene product consists of the first 102 amino acids and a truncation of the last 43 amino acids of the paired domain, as well as the linker region, the homeodomain, and the P/S/T region.

One missense mutation has been identified. The Pax6^4Neu mutation results in an amino acid substitution, Ser273 to Pro, in the third helix of the homeodomain. The predicted gene product contains a single amino acid substitution at a site known to be critical for DNA binding activity. Finally, the Pax6^7Neu mutation is a base pair substitution in the Kozak sequence in the 5'-untranslated region of exon 4 and is predicted to affect translation.

We detected two differences between our wild-type Pax6 sequence and the published mouse Pax6 sequence (WALTHER and GRUSS 1991 ) at position 354 (T instead of C) and at position 1263 (G instead of A). The differences do not affect the amino acids specified. Our sequence data are for strain C3H, were based upon analyses of cDNA, and were confirmed by analysis of genomic DNA. All mutant sequences have been submitted to the EMBL DNA sequence database with the following accession numbers: Pax6^2Neu, Y19193; Pax6^3Neu, Y19195; Pax6^4Neu, Y19196; Pax6^5Neu, Y19197; Pax6^6Neu, Y19198; Pax6^7Neu, Y19199; Pax6^9Neu, AJ292077; and Pax6^10Neu, AJ307468.

Heterozygote and homozygote phenotypes of the hypomorph Pax6^4Neu and Pax6^7Neu mutant alleles:
The degree of lens/corneal opacity in mutant heterozygotes and wild-type littermates is given in Table 5. Heterozygous carriers of the Pax6^3Neu truncation mutation express a severe lens/corneal opacity, the median and modal values being 100% opaque. In contrast, carriers of the two presumed hypomorph alleles Pax6^4Neu and Pax6^7Neu express a less severe phenotype. For both mutations the median and modal values were 25% opaque. As a measure of eye size, individual eyes were weighed. Data were first analyzed for the effects of mutant allele (Pax6^3Neu, Pax6^4Neu, or Pax6^7Neu), genotype (heterozygous or wild type), and sex of animal. The effects of both mutant allele (F_{2, 196} = 6.00, P < 0.002) and genotype (F_{1, 196} = 591.07, P < 0.0001) were highly significant. The sex of the animal had no effect on eye weight (F_{1, 196} = 0.12, P 0.729). Data were pooled for sex of animal and the range, mean, and standard error of the mean eye weights for heterozygous carriers and wild-type littermates are given in Table 6. Eye weights of Pax6^3Neu carriers were 78% that of wild-type littermates. The corresponding values for the Pax6^4Neu and Pax6^7Neu alleles were less severe (83 and 86%, respectively). Examination of embryos from inter se crosses of Pax6^4Neu heterozygotes (Fig 2B) revealed that the resultant phenotypes were less severe than that observed for carriers of a Pax6 null allele (HOGAN et al. 1988 ). We further characterized the extent of nasal development, since Pax6 is essential for normal development of this organ (HOGAN et al. 1988 ; GRINDLEY et al. 1995 ; QUINN et al. 1996 ). Previous studies have shown that Pax9 expression marks the mesenchymal domains of the medial and lateral nasal processes after E10.5 of mouse development (PETERS et al. 1998 ). Using the Pax9^lacZ allele as a reporter, we found that the lateral and medial nasal processes were completely absent in homozygotes for a Pax6 null allele, while the lateral and medial nasal processes were present but of reduced size in homozygotes for the Pax6^4Neu allele (Fig 2C). Histological examination indicated that, in comparison to homozygous wild type (Fig 3A), there was hyperplasia of the cartilaginous parts of the nasal anlage and a concomitant reduction in size of the nasal cavities and their associated nasal epithelia in homozygous Pax6^4Neu mutant embryos (Fig 3B).

View larger version (42K): In this window In a new window Download PPT slide   Figure 3. The extent of eye and nasal development in wild-type and homozygous E16 Pax64Neu or Pax67Neu mutant embryos. (A) Homozygous wild type showing a well-developed eye with lens (le) and retina (ret) and normal nasal structure with nasal septum (ns) and nasal cavities (nc). (B) Homozygous Pax64Neu mutant showing premature termination of eye development and abnormal nasal development. The tip of the optic stalk is a considerable distance from the surface ectoderm. At the tip of the optic stalk an enlargement and invagination of the neuroectoderm form a pseudo-optic cup (oc). From the surface ectoderm there is an invagination (inv) that does not make contact with the pseudo-optic cup and a lens does not develop. The pseudo-optic cup is surrounded by cartilage (cart), which is an extension of the hyperplastic nasal cartilage. Development of the nasal organ is characterized by a hyperplasia of the cartilage, no nasal septum, and only rudimentary nasal cavities (nc). (C) Enlargement of the pseudo-optic cup region in the homozygous Pax64Neu mutant embryo from B, showing the invaginated neuroectoderm of the tip of the optic stalk (neur), lack of contact between the tip of the optic stalk and the invaginated surface ectoderm (inv), and the cartilaginous structures (cart) surrounding the tip of the enlarged and invaginated optic stalk. (D) Homozygous Pax67Neu mutant embryo. Premature eye development and abnormal nasal development are similar to that observed for the Pax64Neu mutation, although hyperplasia of the nasal cartilage was not observed. (E) Enlargement of the pseudo-optic cup region in the homozygous Pax67Neu mutant embryo from D. The invaginated neuroectoderm of the tip of the optic stalk and a lack of contact between the tip of the optic stalk and the invaginated surface ectoderm were similar to that seen in homozygous Pax64Neu mutant embryos. Pigmented cells (pig) can be seen in the region of the pseudo-optic cup.

The extent of eye development in homozygous mutant Pax6^4Neu and Pax6^7Neu embryos is most interesting. Whereas in homozygous wild type (Fig 3A) eye development is complete, with a well-developed retina and lens, eye development in homozygous mutants terminated prematurely (Fig 3, B–E). The optic stalk is not in contact with the surface ectoderm. However, the tip of the optic stalk is enlarged and invaginated to form a "pseudo-optic cup." In homozygous mutant Pax6^4Neu embryos, the enlarged and invaginated tip of the optic stalk is surrounded by cartilage. Invaginations from the surface ectoderm are evident but remain a considerable distance from the tip of the optic stalk. Lens did not develop. Homozygotes for Pax6 null mutations die shortly after birth. In contrast, homozygous Pax6^4Neu and Pax6^7Neu offspring survive up to 24–36 hr after birth. Together, the above observations suggest Pax6^4Neu and Pax6^7Neu are hypomorph alleles.

Three-dimensional model and DNA-binding activity of the Pax6^4Neu mutant gene product:
The site of the Pax6^4Neu missense mutation, Ser273, is located in the third helix of the homeodomain (position 50 in the homeodomain, position 9 in the third helix). The substitution of Pro for Ser is predicted to interrupt the structure of the third helix (Fig 4), which, in three-dimensional structures of homologous homeodomain proteins complexed with DNA, inserts into the major groove of the DNA target. We compared the binding activity of the wild type and Pax6^4Neu mutant homeodomains to the palindromic P3 target oligonucleotide, since the Pax6 homeodomain binds specifically to the P3 sequence (CZERNY and BUSSLINGER 1995 ). Comparable concentrations of wild-type and mutant homeodomain proteins were assayed (Fig 5A). The binding activity of the Pax6^4Neu homeodomain to the P3 oligonucleotide was shown to be greatly reduced in comparison to the binding activity of the wild-type homeodomain (Fig 5B). Only very faint traces of monomer protein-DNA complex were observed. The mutant protein competes weakly with the wild-type protein, reducing the level of dimer protein-DNA complex formation.

View larger version (43K): In this window In a new window Download PPT slide   Figure 4. Three-dimensional model of the Pax6 homeodomain (green) docked to the DNA recognition sequence (cyan). The protein is represented as a backbone ribbon. The position of the mutated residue 9 is marked in red. (A) Wild-type homeodomain with the recognition -helix inserted into the DNA major groove. (B) In the mutant Pax64Neu homeodomain the substitution of serine to proline leads to disruption of the recognition -helix such that the DNA major groove cannot be recognized.

View larger version (152K): In this window In a new window Download PPT slide   Figure 5. DNA-binding assay of the wild-type and mutant Pax6 homeodomain. (A) SDS-PAGE and Western blot analysis of the homeodomain of Pax6 wild-type and mutant Pax64Neu alleles expressed in E. coli as fusion proteins. The efficiency of expression was similar for wild-type and mutant alleles as observed by SDS-PAGE (lanes 1–4) and Western blotting (lanes 5–7). To determine the amounts of protein, Coomassie staining intensities of 3 µl of 1:10 and 1:50 crude lysate were compared with known amounts of marker proteins. For the Western blots a Pax6-homeodomain-specific antiserum was used (gift from Dr. S. Saule, Institut Pasteur de Lille). Lane 1, marker; lane 2, vector; lane 3, wild type; lane 4, Pax64Neu; lane 5, wild type; lane 6, Pax64Neu; lane 7, marker. (B) DNA-binding properties of the homeodomain of Pax6 wild-type and mutant Pax64Neu alleles analyzed by electrophoretic mobility shift assay. The target oligonucleotide was P3. Whole cell extracts from the transformed E. coli strain M15pREP4 were incubated with 15 fmol of the digoxygenin-labeled P3 oligonucleotide. Binding specificity was demonstrated by competition of the labeled olignucleotide with an excess of unlabeled oligonucleotide and by a supershift assay with Pax6 homeodomain-specific antibody. Lane 1, oligonucleotide alone; lanes 2–4, oligonucleotide with the wild-type Pax6 homeodomain (25, 50, or 100 ng, respectively); lane 5, as lane 4 with 1.5 pmol unlabeled oligonucleotide; lanes 6–7, oligonucleotide with the mutant Pax64Neu homeodomain (50 or 100 ng, respectively); lane 8, oligonucleotide with 100 ng wild type and 100 ng Pax64Neu homeodomain; lane 9, as lane 4 with 2 µl of the Pax6-homeodomain-specific antiserum. The positions of the free oligonucleotide (O), the monomeric (M) and dimeric (D) DNA-protein complexes, and the antibody DNA-protein complex shifted band (S) are marked.

Levels of Pax6 expression in homozygous Pax6^4Neu and Pax6^7Neu mutant mice:
The levels of Pax6 expression in homozygous mutant embryos are given in Fig 6. In comparison to homozygous wild-type embryos, normal levels of Pax6 expression were observed in homozygous Pax6^4Neu embryos. For homozygous Pax6^7Neu mutant embryos, the level of Pax6 expression was greatly reduced. The lack of signal associated with homozygous Pax6^3Neu embryos can be taken only as a negative control for the Western blot since the Pax6^3Neu truncation mutation results in loss of the homeodomain against which the anti-Pax6 antiserum reacts.

View larger version (38K): In this window In a new window Download PPT slide   Figure 6. Levels of Pax6 expression in wild-type and homozygous mutant embryos. Protein extracts from heads of E15 embryos were analyzed by Western blot for Pax6 expression. Estimation of protein concentration, electrophoresis procedures, and Pax6-homeodomain-specific antiserum were as given in Fig 5. Fourteen micrograms of total protein were electrophoresed per lane. Lane 1, Pax6 +/+; lane 2, Pax67Neu -/-; lane 3, Pax64Neu -/-; lane 4, Pax63Neu -/-.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Here we have extended the Pax6 allelic series in the mouse. We have shown by mapping, molecular characterization, and phenotype analyses that seven alleles result in truncated gene products and express phenotypic defects characteristic for null mutations. Most importantly, we have identified two potential hypomorph alleles.

Truncation mutations:
Six Pax6 mutations have been previously described. The Pax6^LacZ (ST-ONGE et al. 1997 ) was constructed and the initiation start site as well as the entire paired box were replaced by the ß-galactosidase reporter gene and the neo-selector gene. Pax6^Sey-Dey is due to an intergenic deletion including the Pax6 and Wt1 genes (GLASER et al. 1990 ). Pax6^Sey is a base pair substitution creating a premature stop codon in the linker region (HILL et al. 1991 ). Pax6^Sey-Neu is a splice site mutation resulting in a predicted gene product truncated from the N-terminal end of the P/S/T region (HILL et al. 1991 ). Pax6^Coop is a base pair substitution that creates a premature stop codon between the homeodomain and the P/S/T region (LYON et al. 2000 ). Pax6^Sey-H is a large deletion and most likely involves additional linked loci (HILL et al. 1991 ). Together with the presently described mutations, an allelic series exists for truncation mutations with the order from the longest gene product to total gene ablation as follows: Pax6^Sey-Neu, Pax6^6Neu, Pax6^Coop, Pax6^2Neu, Pax6^Sey, Pax6^3Neu, Pax6^5Neu, Pax6^10Neu, Pax6^9Neu, and Pax6^LacZ. LYON et al. 2000 have observed very small vestiges of unpigmented eyes in homozygous Pax6^Coop embryos, raising the possibility that this mutation may be a hypomorph allele. Although a direct comparison of the phenotypic effects of the different Pax6 alleles in one laboratory has not been carried out, some appropriate data are available. In homozygous Pax6^Sey and Pax6^Sey-Neu mutants, the optic stalk develops distally, there is a condensation of mesenchymal-like cells underneath the surface ectoderm, and the overlying surface ectoderm appears to form a pit, but neither a lens nor an eye develops (GRINDLEY et al. 1995 ; SCHEDL et al. 1996 ). We observed pits in the presumptive eye region of homozygous embryos for all mutations reported here, including the shortest truncation allele in which none of the Pax6 domains remain intact. Thus, the development of such pits may be a component of the phenotype associated with null alleles. The gross morphology of heterozygous and homozygous carriers of all truncation alleles reported here is similar, which suggests that Pax6 gene function requires all domains to be intact. Destruction of the paired domain or the homeodomain should lead to loss of gene function since each participates directly in DNA binding. The amino acid composition of the P/S/T region of the Pax6 protein resembles the activation domains of the transcription factors Oct-1 and Oct-2 (TANAKA and HERR 1990 ). Such activation domains are important for the interaction of DNA-bound transcription factors with other proteins required for transcription.

Potential hypomorph alleles:
In homozygous Pax6 null mutants the optic vesicle makes initial contact with the surface ectoderm but their contact is subsequently lost. The tip of the optic vesicle is broader but does not invaginate in the early steps of optic cup formation and the surface ectoderm does not thicken or invaginate, a process that normally marks the early steps of lens formation (HOGAN et al. 1986 ; SCHMAHL et al. 1993 ; GRINDLEY et al. 1995 ). Studies utilizing chimeric mice have demonstrated that Pax6 activity is essential for surface ectoderm cells to participate in lens formation (QUINN et al. 1996 ). Our preliminary description of homozygous Pax6^4Neu and Pax6^7Neu mutants indicates that eye development proceeds further than that observed for homozygous null mutations. Both the tip of the optic vesicle and the surface ectoderm invaginate. However, contact is not maintained and neither a lens nor a true optic cup develops.

The most conserved segment of the 300 homeodomains known to date is helix III, which makes extensive DNA contacts in the major groove of the DNA target sequence (GEHRING et al. 1994 ). In all Pax6 homologs, residue 9 of the third helix is serine. Mutagenesis studies in vitro have confirmed that the amino acid at position 9 of the third helix is a major determinant for the DNA-binding specificity of the homeodomain (HANES and BRENT 1989 ; TREISMAN et al. 1989 ). The interaction of the N-terminal end of the Pax6 homeodomain and the minor groove of the P3 target sequence has been observed in structural studies (GEHRING et al. 1994 ; QIAN et al. 1994B ). Cooperative dimerization is a two-step process that cannot take place without an efficient monomer binding event (WILSON et al. 1993 ). Thus our observation of faint traces of Pax6^4Neu mutant monomer protein-DNA complex may be due to the weak interaction of the intact N-terminal end of the homeodomain with the minor groove of the target DNA.

The Pax6^7Neu allele is an A-to-T transversion in the 5'-untranslated mRNA region, at position -3 of the Kozak sequence that surrounds the translation start codon. From a survey of 699 vertebrate mRNAs for the initiation site, CC(A/G)CCaugG emerged as the consensus sequence (KOZAK 1987A ). Within this consensus motif the purine at position -3 is the most highly conserved in all eukaryotic mRNAs and mutations at this position affect translation more profoundly than point mutations at other positions (KOZAK 1986 ). In the absence of a purine at position -3, however, a G residue following the AUG codon (position +4) is essential for efficient translation (KOZAK 1987B ).

The sequence surrounding the AUG translation start site, CCAGCaugC, (WALTHER and GRUSS 1991 ) in the wild-type Pax6 gene agrees well with the Kozak consensus sequence. The transversion of A to T at position -3 of the Kozak sequence in Pax6^7Neu combined with C at position +4 suggests that the efficiency of translation is reduced. The reduced levels of Pax6 product in the Western blot analysis confirmed this prediction. Our observation of homozygous mutant embryos with rudimentary eyes suggests that the mutation Pax6^7Neu retains some Pax6 rest activity and may represent a hypomorph allele.

Human PAX6 mutation database:
Human patients expressing aniridia, Peters' anomaly, congenital cataract, keratitis, or foveal hypoplasia have been shown to be heterozygous carriers of PAX6 mutations (GLASER et al. 1992 , GLASER et al. 1994 ; JORDAN et al. 1992 ; EPSTEIN et al. 1994 ; HANSON et al. 1994 ; MIRZAYANS et al. 1995 ; AZUMA et al. 1996 ). In addition, one patient expressing anophthalmia and CNS defects has been shown to be a compound heterozygote for PAX6 mutations (GLASER et al. 1994 ). A database of human PAX6 mutations has been created (BROWN et al. 1998 ) and recently reviewed (http://www.hgu.mrc.ac.uk/Softdata/PAX6). A total of 201 mutations were considered, of which 143 were base pair substitutions, 56 were small deletions, insertions, or deletion/insertions, one was a large deletion, and one was a CA repeat in an intron. The majority of mutations were predicted to result in premature termination of translation and 38 missense mutations were recovered. Most of the recovered missense mutations expressed total aniridia similar to truncation alleles. However, some missense mutations expressed an intermediate phenotype. The mutation R26G (Arg to Gly) was described as a hypomorph allele (TANG et al. 1997 ) and the mutation A33P (Ala to Pro) was associated with a partial aniridia with significant iris remnants (HANSON et al. 1999 ). Both mutations are in the N terminus of the paired domain, which is involved in the binding to the paired domain target DNA. The missense mutations A79E (Ala to Glu) in the paired domain and R208Q (Arg to Gln) in the linker region express a milder phenotype (GRONSKOV et al. 1999 ). Only one missense mutation in the homeobox was detected, Q255H (Gln to His), at a highly conserved site in the recognition helix. The mutation is likely a hypomorph allele since it is associated with sporadic aniridia (CHAO et al. 2000 ). The value of an extensive allelic series of mouse Pax6 mutations is that it provides animal models with which to study in detail gene function in development. The hypomorph mutant alleles may be of special interest for studying the effect of a Pax6 gene product dose on the resulting phenotypes that develop. Our continued efforts to characterize dominant eye mutations recovered in mutagenesis experiments in the mouse should lead to the identification of additional Pax6 mutant alleles.

	FOOTNOTES

¹ Present address: University of Newcastle upon Tyne, Institute of Human Genetics, Newcastle upon Tyne, NE1 7RU United Kingdom.

	ACKNOWLEDGMENTS

We thank Bianca Hildebrand and Irmgard Zaus for their expert technical assistance; Utz Linzner, Institute of Pathology/BIODV Workgroup, for the synthesis of the oligonucleotide primers; and Professor S. Saule for the gift of the rabbit anti-mouse-Pax6-homeodomain. The Tskgel DNA-NPR-S ion exchange column for the HPLC purification of PCR products was kindly provided by TosoHaas, Stuttgart, Germany. Research was supported in part by contract no. CHRX-CT93-0181 from the Commission of the European Communities and National Institutes of Health grant R01EY10321.

Manuscript received April 27, 2001; Accepted for publication September 28, 2001.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AZUMA, N., S. NISHINA, H. YANAGISAWA, T. OKUYAMA, and M. YAMADA, 1996  PAX6 missense mutation in isolated foveal hypoplasia. Nat. Genet. 13:141-142[Medline].

BENTLEY, C. A., M. P. ZIDEHSARAI, J. C. GRINDLEY, A. F. PARLOW, and S. BARTH-HALL et al., 1999  Pax6 is implicated in murine pituitary endocrine function. Endocrine 10:171-177[Medline].

BROWN, A., M. MCKIE, V. VAN HEYNINGEN, and J. PROSSER, 1998  The human PAX6 mutation database. Nucleic Acids Res. 26:259-264[Abstract/Free Full Text].

CHAO, L. Y., V. HUFF, L. C. STRONG, and G. F. SAUNDERS, 2000  Mutation in the PAX6 gene in twenty patients with aniridia. Hum. Mutat. 15:332-339[Medline].

CZERNY, T. and M. BUSSLINGER, 1995  DNA-binding and transactivation properties of Pax-6: three amino acids in the paired domain are responsible for the different sequence recognition of Pax-6 and BSAP (Pax-5). Mol. Cell. Biol. 15:2858-2871[Abstract].

EPSTEIN, J. A., T. GLASER, J. CAI, L. JEPEAL, and D. S. WALTON et al., 1994  Two independent and interactive DNA-binding subdomains of the Pax6 paired domain are regulated by alternative splicing. Genes Dev. 8:2022-2034[Abstract/Free Full Text].

FAVOR, J., 1983  A comparison of the dominant cataract and recessive specific-locus mutation rates induced by treatment of male mice with ethylnitrosourea. Mutat. Res. 110:367-382[Medline].

FAVOR, J., 1986  The frequency of dominant cataract and recessive specific-locus mutations in mice derived from 80 and 160 mg ethylnitrosourea per kg body weight treated spermatogonia. Mutat. Res. 162:69-80[Medline].

FAVOR, J. and A. NEUHÄUSER-KLAUS, 1994  Genetic mosaicism in the house mouse. Annu. Rev. Genet. 28:27-47[Medline].

FAVOR, J. and A. NEUHÄUSER-KLAUS, 2000  Saturation mutagenesis for dominant eye morphological defects in the mouse Mus musculus. Mamm. Genome 11:520-525[Medline].

FAVOR, J., A. NEUHÄUSER-KLAUS, and U. H. EHLING, 1990  The frequency of dominant cataract and recessive specific-locus mutations and mutation mosaics in F₁ mice derived from post-spermatogonial treatment with ethylnitrosourea. Mutat. Res. 229:105-114[Medline].

FAVOR, J., P. GRIMES, A. NEUHÄUSER-KLAUS, W. PRETSCH, and D. STAMBOLIAN, 1997  The mouse Cat4 locus maps to Chromosome 8 and mutants express lens-corneal adhesion. Mamm. Genome 8:403-406[Medline].

GEHRING, W. J., Y. Q. QIAN, M. BILLETER, K. FURUKUBO-TOKUNAGA, and A. F. SCHIER et al., 1994  Homeodomain-DNA recognition. Cell 78:211-223[Medline].

GLASER, T., J. LANE, and D. HOUSMAN, 1990  A mouse model of the aniridia-Wilms tumor deletion syndrome. Science 250:823-827[Abstract/Free Full Text].

GLASER, T., D. S. WALTON, and R. L. MAAS, 1992  Genomic structure, evolutionary conservation and aniridia mutations in the human PAX6 gene. Nat. Genet. 2:232-239[Medline].

GLASER, T., L. JEPEAL, J. G. EDWARDS, S. R. YOUNG, and J. FAVOR et al., 1994  PAX6 gene dosage effect in a family with congenital cataracts, aniridia, anophthalmia and central nervous system defects. Nat. Genet. 7:463-471[Medline].

GOSSLER, A., and J. ZACHGO, 1993 Gene and enhancer trap screens in ES cell chimeras, pp. 181–227 in Gene Targeting: A Practical Approach, edited by A. L. JOYNER. Oxford University Press, New York.

GÖTZ, M., A. STOYKOVA, and P. GRUSS, 1998  Pax6 controls radial glia differentiation in the cerebral cortex. Neuron 21:1031-1044[Medline].

GRINDLEY, J. C., D. R. DAVIDSON, and R. E. HILL, 1995  The role of Pax-6 in eye and nasal development. Development 121:1433-1442[Abstract].

GRINDLEY, J. C., L. K. HARGETT, R. E. HILL, A. ROSS, and B. L. HOGAN, 1997  Disruption of Pax6 function in mice homozygous for the Pax6^Sey-1Neu mutation produces abnormalities in the early development and regionalization of the diencephalon. Mech. Dev. 64:111-126[Medline].

GRØNSKOV, K., T. ROSENBERG, A. SAND, and K. BRØNDUM-NIELSEN, 1999  Mutational analysis of PAX6: 16 novel mutations including 5 missense mutations with a mild aniridia phenotype. Eur. J. Hum. Genet. 7:274-286[Medline].

HANES, S. D. and R. BRENT, 1989  DNA specificity of the Bicoid activator protein is determined by homeodomain recognition helix residue 9. Cell 57:1275-1283[Medline].

HANSON, I. M., J. M. FLETCHER, T. JORDAN, A. BROWN, and D. TAYLOR et al., 1994  Mutations at the PAX6 locus are found in heterogeneous anterior segment malformations including Peters' anomaly. Nat. Genet. 6:168-173[Medline].

HANSON, I., A. CHURCHILL, J. LOVE, R. AXTON, and T. MOORE et al., 1999  Missense mutations in the most ancient residues of the PAX6 paired domain underlie a spectrum of human congenital eye malformations. Hum. Mol. Genet. 8:165-172[Abstract/Free Full Text].

HEINZMANN, U., J. FAVOR, J. PLENDL, and G. GREVERS, 1991  Entwicklungsstörung des olfaktorischen Organs. Ein Beitrag zur kausalen Genese bei einer Mausmutante. Verh. Anat. Ges. 85(Anat. Anz. Suppl. 170):511-512.

HILL, R. E., J. FAVOR, B. L. M. HOGAN, C. C. T. TON, and G. F. SAUNDERS et al., 1991  Mouse Small eye results from mutations in a paired-like homeobox-containing gene. Nature 354:522-525[Medline].

HOGAN, B. L. M., G. HORSBURGH, J. COHEN, C. M. HETHERINGTON, and G. FISHER et al., 1986  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. 97:95-110[Medline].

HOGAN, B. L. M., E. M. A. HIRST, G. HORSBURGH, and C. M. HETHERINGTON, 1988  Small eye (Sey): a mouse model for the genetic analysis of craniofacial abnormalities. Development 103(Suppl.):115-119.

HRABÉ DE ANGELIS, M., H. FLASWINKEL, H. FUCHS, B. RATHKOLB, and D. SOEWARTO et al., 2000  Genome-wide, large-scale production of mutant mice by ENU mutagenesis. Nat. Genet. 25:444-447[Medline].

JORDAN, T., I. HANSON, D. ZALETAYEV, S. HODGSON, and J. PROSSER et al., 1992  The human PAX6 gene is mutated in two patients with aniridia.. Nat. Genet. 1:328-332[Medline].

KAUFMAN, M. H., H.-H. CHANG, and J. P. SHAW, 1995  Craniofacial abnormalities in homozygous Small eye (Sey/Sey) embryos and newborn mice. J. Anat. 186:607-617.

KIOUSSI, C., S. O'CONNELL, L. ST-ONGE, M. TREIER, and A. S. GLEIBERMAN et al., 1999  Pax6 is essential for establishing ventral-dorsal cell boundaries in pituitary gland development. Proc. Natl. Acad. Sci. USA 96:14378-14382[Abstract/Free Full Text].

KISSINGER, C. R., B. LIU, E. MARTIN-BLANCO, T. B. KORNBERG, and C. O. PABO, 1990  Crystal structure of an engrailed homeodomain-DNA complex at 2.8 Å resolution: a framework for understanding homeodomain-DNA interactions. Cell 63:579-590[Medline].

KOZAK, M., 1986  Point mutations define a sequence flanking the AUG initiator codon that modulates translation by eukaryotic ribosomes. Cell 44:283-292[Medline].

KOZAK, M., 1987a  An analysis of 5'-noncoding sequences from 699 vertebrate messenger RNAs. Nucleic Acids Res. 15:8125-8148[Abstract/Free Full Text].

KOZAK, M., 1987b  At least six nucleotides preceding the AUG initiator codon enhance translation in mammalian cells. J. Mol. Biol. 196:947-950[Medline].

KRÜGER, K., T. HERMANN, V. ARMBRUSTER, and F. PFEIFER, 1998  The transcriptional activator GvpE for the halobacterial gas vesicle genes resembles a basic region leucine-zipper regulatory protein. J. Mol. Biol. 279:761-771[Medline].

LYON, M. F., D. BOGANI, Y. BOYD, P. GUILLOT, and J. FAVOR, 2000  Further genetic analysis of two autosomal dominant mouse eye defects, Ccw and Pax6^coop. Mol. Vis. 6:199-203[Medline].

MANLY, K. F., 1993  A Macintosh program for storage and analysis of experimental genetic mapping data.. Mamm. Genome 4:303-313[Medline].

MIRZAYANS, F., W. G. PEARCE, I. M. MACDONALD, and M. A. WALTER, 1995  Mutation of the PAX6 gene in patients with autosomal dominant keratitis. Am. J. Hum. Genet. 57:539-548[Medline].

NOLAN, P. M., J. PETERS, M. STRIVENS, D. ROGERS, and J. HAGAN et al., 2000  A systematic, genome-wide, phenotype-driven mutagenesis programme for gene function studies in the mouse. Nat. Genet. 25:440-443[Medline].

PEARSON, W. R., 1990 Rapid and sensitive sequence comparison with FASTP and FASTA, pp. 63–98 in Methods in Enzymology, Vol. 183, edited by R. F. DOOLITTLE. Academic Press, San Diego.

PETERS, H., A. NEUBÜSER, K. KRATOCHWIL, and R. BALLING, 1998  Pax9-deficient mice lack pharyngeal pouch derivatives and teeth and exhibit craniofacial and limb abnormalities. Genes Dev. 12:2735-2747[Abstract/Free Full Text].

PROSSER, J. and V. VAN HEYNINGEN, 1998  PAX6 mutations reviewed. Hum. Mutat. 11:93-108[Medline].

QIAN, Y. Q., K. FURUKUBO-TOKUNAGA, D. RESENDEZ-PEREZ, M. MÜLLER, and W. J. GEHRING et al., 1994a  Nuclear magnetic resonance solution structure of the fushi tarazu homeodomain from Drosophila and comparison with the Antennapedia homeodomain. J. Mol. Biol. 238:333-345[Medline].

QIAN, Y. Q., D. RESENDEZ-PEREZ, W. J. GEHRING, and K. WÜTHRICH, 1994b  The des(1–6)Antennapedia homeodomain: comparison of the NMR solution structure and the DNA-binding affinity with the intact Antennapedia homeodomain. Proc. Natl. Acad. Sci. USA 91:4091-4095[Abstract/Free Full Text].

QUINN, J. C., J. D. WEST, and R. E. HILL, 1996  Multiple functions for Pax6 in mouse eye and nasal development. Genes Dev. 10:435-446[Abstract/Free Full Text].

SCHEDL, A., A. ROSS, M. LEE, D. ENGELKAMP, and P. RASHBASS et al., 1996  Influence of PAX6 gene dosage on development: overexpression causes severe eye abnormalities. Cell 86:71-82[Medline].

SCHMAHL, W., M. KNOEDLSEDER, J. FAVOR, and D. DAVIDSON, 1993  Defects of neuronal migration and the pathogenesis of cortical malformations are associated with Small eye (Sey) in the mouse, a point mutation at the Pax-6-locus. Acta Neuropathol. 86:126-135