Genetic and Allelic Heterogeneity of Cryg Mutations in Eight Distinct Forms of Dominant Cataract in the Mouse

Abstract

purpose. The purpose of this study was the characterization of eight new dominant cataract mutations.

methods. Lenses of mutant mice were described morphologically and histologically. Each mutation was mapped by linkage studies. The candidate genes (the Cryg gene cluster and the closely linked Cryba2 gene) were sequenced.

results. Molecular analysis confirmed all mutations in Cryg genes. Five mutations lead to amino acid exchanges, two are due to premature stop codons, and one is a 10-bp deletion in the Cryge gene. Morphologically, mutant carriers expressed nonsyndromic cataracts, ranging from diffuse lenticular opacities (Crygd ^ENU910 and Cryge ^ENU449 ), to dense nuclear and subcortical opacity (Crygd ^K10 , Crygc ^MNU8 , Cryge ^Z2 , Crygd ^ENU4011 , and Cryge ^ADD15306 ), to dense nuclear opacity and ruptured lenses (Cryga ^ENU469 ). Results of histologic analyses correlate well with the severity of lens opacity, ranging from alterations in the process of secondary fiber nucleus degradation to lens vacuoles, fiber degeneration, and disruption of the lens capsule.

conclusions. In total, 20 mutations have been described that affect the Cryg gene cluster: Nine mutations affect the Cryge gene, but only one affects the Crygb or Crygf genes. No mutation was observed in the closely linked Cryba2. Two mutations occur at the same site in the Crygd and Cryge genes (Leu45→Pro). The unequal distribution of mutations suggests hot spots in the Cryg genes. The overall high number of mutations in these genes demonstrates their central role in the maintenance of lens transparency.

The β- and γ-crystallins were first characterized by Mörner ¹ more than 100 years ago. Based on their unique folding structure, they are now recognized as members of one β/γ-crystallin superfamily. The corresponding genes are expressed preferentially in the eye and mainly in the ocular lens, and low expression is found in the retina, ^{2 3} brain, and testes. ^{4 5} The common characteristic of all β- and γ-crystallins is the Greek key motif, which allows a dense packing of proteins in the ocular lens. The Cryg genes in all mammals consist of three exons: The first one codes for three amino acids, and the subsequent two are responsible for two Greek key motifs each. Biochemically, the γ-crystallins are characterized as monomers with a molecular mass of 21 kDa (for reviews see Refs. ^{6 and 7} ). 

Six members of the Cryg family (Cryga→Crygf) are located in a cluster on mouse chromosome 1 or the long arm of human chromosome 2, region 33-35, whereas the seventh Cryg gene (Crygs) maps to mouse chromosome 16 and human chromosome 3. The Cryba2 gene encoding the βA2-crystallin is located approximately 8 cM distal to the mouse Cryg gene cluster. In humans, the relative map positions of the CRYG gene cluster and the CRYBA2 gene are similar to the CRYBA2 located on the long arm of chromosome 2, region 34-36. ⁸  

In mice, mutations in all six genes of the Cryg gene cluster have been identified and shown to lead to dominant, congenital cataracts: Cryga ^ENU-436 , Crygb ^Nop , ⁹ Crygc ^Chl3 , ¹⁰ Crygd ^Lop12 , ¹¹ and Crygd ^Aey4 . ¹² Five cataract-causing alleles of Cryge have been reported so far in the mouse: Cryge ^Elo , ¹³ Cryge ^t , ⁹ Cryge ^nz , ¹⁴ Cryge ^Aey1 , ¹⁵ and Cryge ^ENU418 . ¹⁶ Moreover, one dominant cataract was reported recently in the sixth gene of the Cryg cluster, Crygf. ¹⁷ In addition, mutations in the mouse Crygs gene were demonstrated to be causative of a dominant ¹⁸ and a recessive cataract. ¹⁹ Several hereditary cataracts in humans have also been shown to be caused by mutations in CRYG genes. ^{20 21 22 23 24}  

In the course of a large-scale mouse mutagenesis program, ^{25 26} we identified more than 200 mutants with different dominant cataracts. As reported previously, ²⁶ the largest subgroup among our collection is located on mouse chromosome 1 close to the Cryg gene cluster and the Cryba2 gene. Looking at this unequal distribution of mutations in the genome, the fundamental question comes up of why this particular region on mouse chromosome 1 is more affected than others. Therefore, it is necessary to characterize, as a first step, the underlying mutations to enable future experiments to provide more detailed analysis of the mechanisms leading to the different types of cataracts. Herein, we report eight novel dominant cataract mutations: Seven come from our own mutagenesis screening and an additional one from a similar study at the Oak Ridge National Laboratory (ORNL, Oak Ridge, TN). Molecular analyses show that all eight mutations affect genes in the Cryg gene cluster on mouse chromosome 1, but none of them affect the closely linked Cryba2 gene. 

Materials and Methods

Animals

For the mutation screens, chemically treated male mice, either (102/ElxC3H/El)F₁ hybrids or the strain DBA/2, were mated with untreated female T-stock mice. In the study at ORNL, mutagenized C3Hf/Sl males were mated with T-stock females. Offspring were ophthalmically examined for eye abnormalities at weaning, with a slit lamp microscope (model SLM30; Carl Zeiss Meditec, Oberkochen, Germany). Presumed mutations from the treatment or control groups or even from the breeding colony were genetically confirmed and further outcrossed to either strain 102/El or (102/ElxC3H/El)F₁ hybrid mice. ²⁷ All mutant lines were subsequently backcrossed to strain C3H. Homozygous mutant lines were established and have been maintained by brother x sister matings. All breeding was performed in the National Center for Environment and Health (GSF, Neuherberg, Germany) animal facility, according to the German Law on the Protection of Animals and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Mutations

Of the mutations characterized in the present study, seven were recovered in Neuherberg. Four were detected in the offspring of mice exposed to ethylnitrosourea (ENU). The founder mutant ENU369 expressed lens vacuolization and total opacity ²⁸ ; ENU449, diffuse total opacity ²⁸ ; ENU910, total cloudy opacity ²⁹ ; and ENU4011, total cloudy opacity (Favor J, unpublished data, 1990). One mutation (MNU8) was recovered in the offspring of mice exposed to methylnitrosourea (MNU), and the founder mutant expressed lens vacuolization and total opacity (Favor J, unpublished data, 1989). The mutant K10 was recovered in the offspring of untreated parental mice and expressed total opacity, whereas the mutant Z2 was discovered in breeding stocks and expressed nuclear and zonular opacity (Favor J, both unpublished, 1985 and 1997). 

One mutation was recovered at ORNL in the offspring of an ENU-treated male. Heterozygous ADD15306 mutants expressed a nuclear and zonular opacity (Selby PB, unpublished data, 1997). 

Mapping of the Mutations

The mutations were mapped relative to microsatellite markers according to methods previously described. ^{30 31} The mutation ADD15306 was mapped commercially (i.puma, Bad Homburg, Germany). Chromosomal positions of genes or markers were taken from the MGI database (http://www.informatics.jax.org; provided in the public domain by Jackson Laboratories, Bar Harbor, ME). Mice were kept under specific pathogen-free conditions at the GSF, according to the German Law on the Protection of Animals. 

Morphologic and Histologic Analysis

For gross documentation, lenses were prepared under a dissecting microscope (MZ APO; Leica, Heidelberg, Germany) and photographed at 25× magnification. For histologic analysis, eyes of 3-week-old homozygous mutant mice were fixed for 24 hours in Carnoy solution, dehydrated, and embedded in plastic medium (JB-4Plus; Polysciences Inc., Eppelheim, Germany) according to the manufacturer’s instructions. Sectioning was performed with an ultramicrotome (Ultratom OMU3; Reichert, Walldorf, Germany). Serial transverse 2-μm sections were cut with a dry glass knife and stained with methylene blue and basic fuchsin. The sections were evaluated by light microscope (Axioplan; Carl Zeiss Meditec, Halbergmoos, Germany). Images were imported into image-processing programs (Photoshop 6.0, Illustrator 9.0; Adobe, Unterschleissheim, Germany). All wild-type controls were strain C3H/El. 

Isolation of RNA, DNA, and PCR Conditions

Genomic DNA was prepared from tail tip, liver, or spleen, and RNA was isolated from lenses (stored at −80°C) of newborn mice, according to standard procedures. cDNA synthesis and PCR for mouse Cryg or Cryba2 genes using genomic DNA or cDNA as template were performed, as described previously. ^{9 15} For amplification of a smaller genomic fragment of Crygc, exon 3, we used an additional internal left-side primer (CCTCAGTGAGGTGCGCTCGC) together with the already-described right-side primer, and the resultant PCR fragment of 279 bp was digested by the restriction enzyme SduI. 

PCR products were sequenced commercially (SequiServe, Vaterstetten, Germany) after cloning into the pCR II vector (Invitrogen, Leek, The Netherlands) or directly after elution from the agarose gel, using kits from Qiagen (Hilden, Germany) or Bio-Rad (Munich, Germany) and subsequent precipitation by ethanol and glycogen. For sequence notation, the A of the ATG start codon in the cDNA was assigned nucleotide position 1. Correspondingly, at the protein level, the first Met was assigned position 1 of the amino acid sequence. 

Computer-Assisted Prediction of the Biochemical Properties of the Mutated Proteins

The analyses were performed using the Proteomics tools of the ExPASy Molecular Biology server (http://www.expasy.ch; provided in the public domain by the Swiss Institute of Bioinformatics, Geneva, Switzerland). In particular, we used Kyte-Doolittle algorithms for hydrophobicity, ³² the TMpred and TopPred2 programs to detect transmembrane domains, GOR4 ³³ for secondary structures, and the ScanProsite program for additional biochemical features. Protein models were calculated using the ExPASy first-view program and RasMol 2.6. 

General

Chemicals were from Merck (Darmstadt, Germany) or Sigma-Aldrich (Deisenhofen, Germany). If not indicated otherwise, the enzymes used for cloning and reverse transcription were from Roche Diagnostics (Mannheim, Germany), and restriction enzymes were from MBI Fermentas (St. Leon-Rot, Germany). 

Results

Morphology and Histology

The lens opacities expressed by eight heterozygous and homozygous mutants are documented in Figure 1 . Compared with the homozygous wild-type (Fig. 1A) , there was a wide spectrum of lens opacities expressed by carriers of the different mutants. Most extreme is the phenotype observed in heterozygotes and homozygotes of the mutation ENU369 (Figs. 1B 1C , respectively), in which dense nuclear opacity was observed, and lenses were ruptured at 3 weeks of age. Heterozygous and homozygous mutants K10 (Figs. 1D 1E) , MNU8 (Figs. 1F 1G) , Z2 (Figs. 1H 1I) , ENU4011 (Fig. 1J 1K) , and ADD15306 (Figs. 1L 1M) all expressed cataracts: a dense nuclear and subcortical cataract in heterozygotes and the more severe opacity in homozygotes. The phenotypes associated with the mutations ENU910 (Figs. 1N 1O) and ENU449 (Figs. 1P 1Q) were, by comparison, much milder and were characterized as diffuse lenticular opacities. 

Histologic analysis demonstrated also a broad spectrum of severity (Fig. 2) . Compared with the wild-type C3H lenses (Fig. 2A) , the mutants ENU910 (Fig. 2H) and ENU449 (Fig. 2I) showed only subtle changes as a darkly stained homogenous mass in the core of the lens. Because methylene blue can also be used as a stain for cell nuclei, the darkly stained areas in these two mutants could be interpreted as homogenously dispersed nuclei acids, liberated from imperfectly degraded cell nuclei. These minor changes correlate with the weak phenotype observed in the slit lamp as well as in isolated lenses. The mutant ADD15306 (Fig. 2G) also showed sections without major abnormalities. In addition to the darkly stained lens core, tiny black spots were present inside the fiber cells, most likely representing fiber cell nuclei, which are only partly degraded. Correspondingly, the wave of the cell nuclei in these three mutants was not as well ordered as in the wild type, indicating also a disturbance of the lens fiber cell differentiation. 

The mutants MNU8 (Fig. 2D) , Z2 (Fig. 2E) , and ENU4011 (Fig. 2F) exhibited an intermediate phenotype. Small vacuoles and swollen fiber cells were present in various parts of the lens as were alterations in the lens core (anteriorly shifted lens core in ENU4011 and Z2), which are consistent with the nuclear and subcortical cataracts observed in the corresponding lenses. In the MNU8 and Z2 mutants, the lens nucleus seemed to be separated from the outer cortex. In particular, in the Z2 mutants, a plaque of epithelial cells was present at the anterior pole. In general, the denucleation process was abnormal, as reported several times in other cataract mutants. ^{34 35}  

The most severe phenotypes were observed in the sections from the ENU369 (Fig. 2B) and K10 (Fig. 2C) mutant mice. In particular, the lens nucleus of the K10 mutant was completely shifted to the anterior epithelium and highly vacuolated. The primary lens fiber cells had degenerated. The secondary fiber cells did not enlarge, leading to a very small lens. In the ENU369 mutant, the posterior lens capsule was ruptured, and lens material was, therefore, also present in the vitreous. Other consequences of the rupture of the posterior lens capsule and the loss of lens material are the high numbers of large vacuoles in the remaining lens. However, in all these cases, the cell nuclei were not fully degraded and could be stained by corresponding techniques. In contrast to the lens, the retinas, iris, cornea were not affected. 

Mapping

The mutation ENU369 has been mapped to chromosome 1 relative to the visible markers fz and ln, and the results suggested that it is located on chromosome 1 at cM 28. ³⁶ We mapped the remaining seven mutations to approximately the same chromosomal region (Table 1) . Therefore, for all mutations, the Cryg genes are considered candidate genes and were sequenced either at the cDNA and/or at the genomic DNA level. Because the Cryba2 gene is very close to the Cryg gene cluster (map distance just 8.8 cM according to the Chromosome Committee Report, 2000; available online at http://www.informatics.jax.org/ccr/searches/index.cgi?year=2000), it was also sequenced. 

Sequence Analysis and Molecular Modeling

In the present study, we identified causative mutations in the Cryga, Crygc, Crygd, and Cryge genes, but none in the Crygb, Crygf, or Cryba2 genes. A schematic overview of the general features of the Cryg genes and the encoded Greek key motifs is given in Figure 3A ; the three-dimensional (3-D) structure of a typical wild-type γ-crystallin is based on crystallography data of the γE-crystallin from the rat and of the bovine γF-crystallin (Fig. 4A) . The mutations will be discussed in alphabetical order (Cryga–Cryge). 

Sequencing of the mutation ENU369 revealed a specific T→C transition at position 127 in exon 2 of Cryga (Fig. 3B) . The mutation creates a new HpaII restriction site that was not found in six different wild-type strains (BALB/c, C57BL/6, JF1, 129, DBA, or C3H), but was present in all four homozygous mutants that were analyzed (Fig. 5A) . The mutant allele was accordingly designated Cryga ^ENU369 . 

At the protein level, it leads to an amino acid change from an aromatic Trp to a basic Arg at position 43 (W43R) in the second Greek key motif (Fig. 4B) . The Trp occurs at this position in all γ-crystallins from mouse, rat, bovine, and human and is essential for the formation of the second Greek key motif. The program ScanProsite predicts that the second Greek key motif will not be formed in the mutant γA-crystallin. Moreover, it predicts that the hydrophobicity of the corresponding region will be changed as well as the isoelectric point (pI) of the entire protein (from pH 7.5 in the wild-type γA-crystallin to pH 8.5 in the mutated form). In addition, ScanProsite suggests that a novel tyrosine kinase phosphorylation site may be created at amino acid position 43-51. 

A mutation in the Crygc gene is causative of the cataractous phenotype in the MNU8 mutant line. The exchange of the regular G with an A at position 471 of the Crygc gene leads to a stop codon and a truncation of the protein at position 157 (Trp157Stop; Fig. 3B ). It destroys one (of two very close) SduI restriction site, which was confirmed by the analysis of four wild-type mice from different genetic backgrounds and four mutants (Fig. 5B) . The observed pattern of bands in the PAGE unexpectedly did not fit the predicted sizes of the DNA fragments. In wild type, the 125- and 133-bp fragments should be separated electrophoretically. Likewise, in mutants, the 146- and 133-bp fragments should be separable. However, sequence analysis of the two wild-type fragments revealed that the upper band is a mixture of the 125- and 133-bp fragments, whereas the lower band represents the 125-bp fragment. Obviously, the mixture of both fragments results in a slower velocity during electrophoresis most likely due to hyperstructures. The sequence analysis of the band in the mutants revealed a mixture of the 146- and 133-bp fragments, consistent with the destruction of the first SduI restriction site. Therefore, we propose the allele symbol Crygc ^MNU8 . 

The ScanProsite program suggests that two biochemically active sites are missing in the truncated γC-crystallin, namely an N-myristoylation site (at amino acid position 158-163) and a PKC phosphorylation site (at amino acid position 166-168); the pI of the protein is calculated to be at pH 6.9 instead of pH 7.6, as in the wild-type allele. Three-dimensional modeling (Fig. 4C) demonstrates that an essential part of the fourth Greek key motif (the third antiparallel strand and a part of the α helix) is missing; however, ScanProsite suggests that the fourth Greek key motif will form. 

Three of the new mutants evaluated in this study are associated with mutations in the Crygd gene: In the ENU4011 mutant line, a T→C exchange was observed at position 134 of the Crygd gene. Because the mutation did not affect a restriction site, it was confirmed at the genomic DNA level by sequencing exons 1 and 2 and their flanking regions in five mutants and by comparison to the database (NM_007776) and to four different wild-type strains (C57BL/6, JF1, DBA, and T-stock). The C at cDNA position 134 was found only in the five homozygous cataractous mutants. Therefore, this new Crygd allele is referred to as Crygd ^ENU4011 . 

The mutation leads to a replacement of Leu at position 45 by Pro (Fig. 3B) . The Leu residue is present in most of the γ-crystallins analyzed so far. It is replaced by Ile in the human γB- and the bovine γD-crystallin and by Val in the bovine γA-crystallin. Codon 45 is located in the second Greek key motif (Fig. 5B) . The ScanProsite program suggests that the folding properties of the γD-crystallin are strongly affected by the substitution of Leu by Pro and that the second Greek key motif will be prevented from forming. 

Two polymorphic sites in this mutant line ENU4011 are shared with the Test stock, one of which is close to the characterized mutation site. Therefore, it might be speculated that this mutation was not induced by ENU, but occurred spontaneously in the germ cells of the untreated T stock mouse (in our standard protocol, all parental mice were ophthalmologically examined for the presence of pre-existing mutations ²⁸ ). 

The second mutation affecting the Crygd gene was recovered in the mutant line ENU910 as an A→T substitution at position 268 (Fig. 3B) . The mutation destroys a Bsp143I restriction site in the Crygd gene. A digest of the corresponding genomic fragment confirmed the absence of the Bsp143I restriction site in all five homozygous mutant mice, but demonstrated its presence in all four wild-type strains tested (Fig. 5C) . Therefore, this allele was named Crygd ^ENU910 . 

At the protein level, this mutation changes the Ile at position 90 to a Phe. Comparing all available data on γD-crystallin amino acid sequences in mammals, only Lys or Met is present at this position without effects on the γ-crystallin functions. Three-dimensional modeling (Fig. 4B) shows that this position is at the beginning of the third Greek key motif, which will form correctly. The pI is slightly higher for the mutated γD-crystallin (pH 7.0 versus pH 6.6 for the wild-type protein). 

The third mutation was confirmed in the K10 mutant line as a unique 432C→G base-pair substitution (Fig. 3B) that destroys an ScaI restriction site. A digest of the corresponding genomic fragment confirmed the absence or presence of the ScaI restriction site within the mutant lines or wild-type strains, respectively (Fig. 5D) . Therefore, this allele was referred to as Crygd ^K10 . 

At the protein level, this mutation creates a stop codon in the linker region between the third and fourth Greek key motif and leads to a truncation of the protein after 143 amino acids. The 3-D model of the truncated protein (Fig. 4D) as well as ScanProsite show that the mutant gene product lacks the last Greek key motif. The tyrosine kinase phosphorylation site (amino acids 147-154), the N-myristoylation site (amino acid position 158-163) and the PKC phosphorylation site (amino acid position 166-168) are not present. The pI of the truncated γD-crystallin is lower (pH 5.9) than the pI of the wild-type protein (pH 6.6). 

Three mutations were detected in the Cryge gene: The spontaneous mutation Z2 exhibits a 10-bp deletion in the very beginning of the Cryge cDNA in exon 2 (del12-21; Fig. 3B ). The mutation creates a new HphI restriction site that was confirmed by analysis of four homozygous mutants and wild-type mice of four different strains (Fig. 5E) . Therefore, the mutation is called Cryge ^Z2 . The deletion of 10 bp creates a new open reading frame consisting of the first three N-terminal amino acids of the Cryge followed by 119 novel amino acids. Computer-based analysis of the mutant protein predicts two transmembrane domains from amino acids 23 or 28 to position 44 and from position 46 or 50 to 68 or 70, and the N and C termini are likely to be at the cytosolic site. ScanProsite suggests also the presence of a threonine-rich region. However, this prediction is below the accepted threshold level and may be spurious. 

The mutation ADD15306 is characterized by an exchange of a T for a C at position 134 (Fig. 3B) . Because this mutation does not affect a restriction site, it was confirmed by sequencing in three additional mutant animals using genomic DNA as substrate, but was never found in wild-type mice of different genetic origins (C57BL/6, T-stock, BALB/c, 129, and JF1). Therefore, this Cryge allele is referred to as Cryge ^ADD15306 . 

The mutation is predicted to lead to a Leu→Pro exchange at codon 45 in the second Greek key motif, the same position as in the mutant Crygd ^ENU4011 (Fig. 4B) , and, similarly, ScanProsite predicts the absence of this particular motif. The pI is reduced in the mutant γE-crystallin (pH 7.1) compared with the wild type (pH 7.7); other alterations are not predicted. 

The third mutation in the Cryge gene was found in the ENU449 mutant line. The mutation is characterized by a G→A exchange at position 376 in exon 3 (Fig. 3B) . It destroys an Eco72I restriction site, which was confirmed in the genomic DNA of five mutants and five wild-type mice of different genetic origin (Fig. 5F) . Therefore, the mutation was named Cryge ^ENU449 . 

It is predicted that the amino acid sequence is affected at position 126 (Val→Met). ScanProsite does not suggest any effect on the Greek key motif formation, because the changed amino acid is considered to be localized just at the beginning of the fourth motif (Fig. 4B) . However, comparing all available protein sequences revealed that the Val residue is highly conserved at this position and present in all γ-crystallins analyzed so far. As for the ADD15306 mutation, the pI of the γE-crystallin affected by the ENU449 mutation is lowered also to pH 7.1. 

Polymorphisms

Several polymorphic sites were identified in the Cryg gene cluster (Table 2) . Most of them occurred in the T-stock mice, but some also in the mutant lines. They will be presented in alphabetical order from Cryga to Crygf. 

In Cryga, all mutant lines tested as well as the wild-type strain T-stock demonstrated a deviation from the database sequence (accession no. NM_007774) at position 201 without effect on the amino acid composition. 

In the Crygb gene, two polymorphic sites were observed. At position 238, the T-stock mice exhibit a T instead of a C, which leads to an exchange of an Arg for a Cys at codon 80. This Cys is also present in human γB- and γC-crystallins and in bovine γC- and γE-crystallins. In the rat γF-crystallin, a His is present at this position. At position 524, a T was present in the T-stock mice, changing the Tyr at codon 175 into a Phe, which is present in rat Crygf at the corresponding position. All other Cryg genes exhibit either Tyr or Leu. 

In the Crygc gene, we identified three polymorphic sites. In all cases, the encoded amino acids are not changed. 

All four polymorphic sites identified in the Crygd gene are predicted to cause changes in the amino acid composition. At position 95, the database (accession no. NM_007776) reported an A, which is replaced in all our sequences by a G, leading to an exchange of His for Arg at codon 32. His at this position is reported also in all other species. In contrast, at position 154/156 the commonly used sequence GCA (Ala) is replaced only in the T-stock by ACG (Thr; nucleotide 52). This position does not seem to be highly conserved among mammals (human, Ser; rat, Thr; bovine, Leu). At position 303/304, the database reports an AG, but all our lines have a GA, which leads to a Val→Met exchange at amino acid position 102. The Met residue is present in most of the other γ-crystallins at this position; a Val was reported for the human γD-crystallin as well as for mouse and rat γA-crystallin. In the bovine γC- and γE-crystallins, an Ile occurs at this position. At nucleotide 488, the database A is replaced in all cases by a G; therefore, the DNA polymorphism causes an exchange from Lys to Arg at codon 163. The Arg has been reported for human and rat γD-crystallin, too. 

During this study, we also identified a small deletion in intron A of the JF1 wild-type mice. The sequence GCCTT is present three times in three wild-type strains (C57BL/6; DBA/2, T-stock), but only twice in the JF1 strain and in the database entry based on C3H mice (accession no. AJ224342). That in the JF1 mice the correct cDNA for Crygd was also observed suggests that this deletion of one of the repeated GCCTT elements does not affect the splicing mechanism in the small first intron of the JF1 Crygd gene. 

Next to the Crygd gene, the Cryge gene showed the most polymorphic sites in the mouse strains investigated. There were several polymorphisms that are present only in a few of the mutant lines or the T-stock strain. Some of them have no effect on the amino acid composition. At position 47, Glu is the predominant form in all other γ-crystallins. At codon 52, a T52A polymorphism is observed in the T-stock mice. It should be noted that, in the γD-crystallin, the common amino acid at this position is also Ala. 

During this study, a further polymorphic site was observed in intron A at position 11: a G is present in the ADD15306 mutant line and in the wild-type strain 129, but an A in all other wild-type strains analyzed including the database (accession no. X57855). Moreover, the frequency of the CTCAG repeat at the 3′-end of intron B is lower in the T-stock mice as reported previously for the (101xC3H)F1 hybrids (GenBank accession no. X57855). 

The specific amplification of the Crygf cDNA is not always successful because of its extensive similarity to the Cryge cDNA, which is usually preferentially expressed. Therefore, in all mutant lines the Crygf gene was analyzed at the genomic level by amplifying the coding (exon) regions together with flanking (intron) regions. Three polymorphic sites were recovered in the T-stock mice, two of which have no effect on the amino acid composition. The 493 A→G exchange in the T-stock and the ENU449 mutant line leads to Gly, which is present in all γ-crystallins at this position; in contrast, the Ser in all other mutant lines seems to be the exception. 

In addition, we checked the expressed sequence tag (EST) database entries sharing the key word γ-crystallin for polymorphic sites and a surprisingly high number of additional polymorphic sites were found. Unfortunately, only a few of them could be attributed to strain-specific (mainly C57BL/6J) polymorphisms, because most of the entries do not give information on the particular strain characterized. However, because in some cases, contradictory results were found, it may be necessary to confirm these sequences before making conclusions based on rough data. 

Discussion

In the present paper, we describe the phenotypes and results of molecular characterizations of eight mouse cataract mutations affecting genes in the Cryg gene cluster. Because the mutations segregate with the phenotypes, the evidence is strong that these mutations are responsible for the cataract phenotypes. 

Genotype–Phenotype Considerations

The eight novel Cryg alleles lead to different phenotypes of congenital lens opacifications, ranging from nuclear, to total, to lamellar cataracts. The cataractous eyes are in most cases smaller, and the mutations show a semidominant mode of inheritance. The most severe phenotype among the eight mutants described herein is represented by the Cryga ^ENU369 mutant, which is caused by a mutation affecting the Cryga gene. In the mouse, it was shown recently that the Cryga gene is expressed very early (at embryonic day [E] 12.5) and at a higher level than the Cryge/f genes. ³⁸ The other Cryg genes are expressed only from E14.5 onward. With the exception of the very mild phenotypes caused by the Crygd ^ENU910 and Cryge ^ENU449 mutations, the other phenotypes are consistent with the hypothesis that the severity of the cataracts increases, if the affected Cryg gene is expressed earlier. Thus, one factor to be considered important in the degree of severity of congenital cataract is the time of onset of gene expression during embryonic development. 

A second factor is the degree to which the mutated gene is altered. In addition to these 8 Cryg gene mutations reported in this study, 12 other mouse mutations have been reported to affect the Cryg gene cluster and to lead to cataracts (Table 3) . Most of these 20 Cryg mutations lead to an amino-acid exchange in an important region of the corresponding γ-crystallin, or they express a truncated form of the γ-crystallin with or without new amino acids. For the mutations with extreme cataract phenotype, the mutant proteins were predicted to be altered drastically. Moreover, there are four frameshift mutations (Cryge ^Aey1 , Cryge ^ENU418 , Cryge ^Nz , and Cryge ^Z2 ) leading to chimeric proteins, which include novel amino acid sequences from the point of the frame shift that have no similarity to the γ-crystallins. 

Among the biophysical data available, the changes of the pIs were interesting. Recently, Ueda et al. ³⁹ reported that during ageing, an acidification of the mouse γ-crystallins appears (on average 0.47 pH units). The acidified forms were almost exclusively found in the insoluble fraction. Among the novel Cryg mutations reported in the present study, the Cryga ^ENU369 mutation has the most severe phenotype. However, the affected γA-crystallin is changed into a basic form (pI changed from pH 7.5 to 8.5). Among all the Cryg mutants listed in Table 3 , the most extreme acidic alteration in the pI occurs in the Crygb ^Nop mutant (pH 7.6 to 5.5), and the most extreme basic alteration occurs in the Cryge ^ENU418 (pH 7.7 to 10.1). However, in both cases the phenotype is not as severe as in the Cryga ^ENU369 or the Cryge ^t , suggesting that the pI is not a valuable parameter for genotype–phenotype correlation. 

Posttranslational modifications of the mutated proteins might occur (e.g., phosphorylation) depending on the site and the nature of the mutation. Such modifications have been predicted for Cryga ^ENU369 , Crygc ^MNU8 and Crygd ^K10 . However, biochemical analyses are required to determine if such posttranslational modifications occur. 

Another approach might address the question of how the mutations and the altered proteins interfere with the ongoing process of terminal differentiation in the lens fiber cells. We could show recently that Cryg-mediated cataractogenesis (in Crygb ^nop , Cryge ^Elo , and Cryge ^t ) is associated with the formation of amyloid fibrils in the nuclei of primary lens fiber cells. This interpretation was supported by the observation of recombinant mutant γ-crystallin forming amyloid fibrils also in cell culture, whereas native γ-crystallins are soluble and nonfilamentous. ³⁸ This process is associated with an inhibition of the degradation of the central lens fiber cell nuclei and of the chromosomal DNA. This is most likely caused by an impaired action of a Mg²⁺-dependent DNase, which was shown to correlate well with the severity of three types of cataract investigated (Cryge ^Ns > Crygb ^Nop > Cryge ^t ). ⁴⁰ This interpretation is supported by recent findings that mice deficient in the DLAD gene (coding for a DNase II–like acid DNase or DNase IIβ) are incapable of degrading DNA during lens cell differentiation and develop nuclear cataract. ⁴¹ However, the biochemical mechanisms during cataractogenesis remain to be elaborated. 

Unequal Distribution of Mutations among the Cryg Gene Cluster

The mutations are not evenly divided among the six Cryg genes in the cluster: nine in the Cryge gene, four in the Crygd gene, three in the Crygc gene, two in the Cryga gene, and only one mutation each in the Crygb and Crygf genes. Surprisingly, not a single mutation or polymorphic site was found in the closely linked Cryba2 gene. The six β-crystallin encoding genes are in general less affected than the Cryg genes; there has been one mutation reported in Cryba1 ⁴² and two in Crybb2. ^{5 43} In this context, it might be interesting to note that two mouse mutations affect also the Crygs gene at mouse chromosome 16 ^{18 19} ; one of them is inherited in a recessive mode even if the mutation leads to a late truncation of the protein (Trp163Stop). ¹⁹ The relatively high rates of mutation induction compared with other genes that cause cataracts, the high number of polymorphic sites, and the unequal distribution among the Cryg gene cluster were unexpected. Also unexpected were the findings of identical mutations in two different genes. The T134C mutation was found in both Crygd (ENU4011) and Cryge (ADD15306). Both were induced by ENU, and it is also of interest that their phenotypes are different. Moreover, the G470A mutation was found both in the mouse Crygd and human CRYGD. ^{11 24}  

Corresponding to the increasing number of characterized cataract mutants in mice, mutations in human CRYG genes have been shown to be associated with cataract formation: the Coppock-like cataract ²² and the variable zonular pulverulent cataract ²¹ with the CRYGC gene and the aculeiform cataract, ²² a punctate cataract, ²³ and a crystal-deposition cataract ²⁰ with mutations in the CRYGD gene. Three other hereditary congenital cataracts with lamellar or nuclear opacity are also associated with mutations and affect the CRYGC or CRYGD genes. ²⁴ No mutation has been reported to date in the CRYGA or CRYGB genes or in the closely linked CRYBA2 gene. 

The only described change in the human CRYGA is an insertion at pos. 43 leading to a frame shift and a premature stop codon after 7 novel amino acids; it is not known whether the corresponding 20-amino acid peptide is stable. In the heterozygous situation this insertion is without pathological consequences; it is not known whether it might lead to cataracts in homozygotes. ⁴⁴ Among the β-crystallin encoding genes, the CRYBB2 is most often affected because of possible recombinations with its closely linked pseudogene ^{45 46 47} ; further mutations have been reported only in CRYBA1 ⁴⁸ and CRYBB1. ⁴⁹  

In conclusion, it is suggested even from the relatively small number of characterized mutations that the 7 Cryg genes are more frequently affected than the 6 Cryb genes. Moreover, even among the Cryg genes there is an unequal distribution of mutations suggesting a sub-group of Cryg genes, which are a particular hot spot for cataract-causing mutations in mouse and man. 

 Submitted for publication July 30, 2003; revised September 11, 2003; accepted October 5, 2003.

 Disclosure: J. Graw, None; A. Neuhäuser-Klaus, None; N. Klopp, None; P.B. Selby, None; J. Löster, None; J. Favor, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Lens opacities expressed in 3-week-old heterozygous and homozygous mutants: (A) C3H wild type; ENU369 (B) heterozygote and (C) homozygote; K10 (D) heterozygote and (E) homozygote; MNU8 (F) heterozygote and (G) homozygote; Z2 (H) heterozygote and (I) homozygote; ENU4011 (J) heterozygote and (K) homozygote; ADD15306 (L) heterozygote and (M) homozygote; ENU910 (N) heterozygote and (O) homozygote; and ENU449 (P) heterozygote and (Q) homozygote.

Figure 2.

 

Histologic analysis of cataractous lenses from homozygous mice. A histologic overview of the entire lens is presented in the top rows of (A) to (H) and a higher magnification of the boxed region is shown in the bottom rows. For the weak phenotypes ENU910 and ENU449, only the overview is demonstrated. (A) C3H, (B) ENU369, (C) K10, (D) MNU8, (E) Z2, (F) ENU4011, (G) ADD15306, (H) ENU910, and (I) ENU449. Bars, 100 μm.

Table 1.

 

Mapping Results from Seven Dominant-Eye Mutants of the Mouse, Relative to Chromosome 1 Microsatellite Markers

ENU449 - (0/102) - D1Mit156 - (2/102) - D1Mit181

D1Mit231 - (31/100) - ENU910 - (1/100) - D1Mit156 - (3/100) - D1Mit181

ENU4011 - (0/113) - D1Mit156 - (3/113) - D1Mit181

D1Mit213 - (9/109) - MNU8 - (0/109) - D1Mit156 - (5/109) - D1Mit181 - (6/109) D1Mit334

Z2 - (1/115) - D1Mit156 - (4/115) - D1Mit181

K10 - (0/107) - D1Mit156 - (6/107) - D1Mit181

D1Mit211 - (5/38) - ADD15306 - (2/38) - D1Mit216 - (14/38) - D1Mit206

 Numbers in parentheses indicate the number of recombinants observed (numerator) in the total number of offspring genotyped (denominator) between the loci.

Figure 3.

 

A schematic overview of the Cryg genes, the encoded mutations, and the γ-crystallin structure. (A) A typical Cryg gene consisting of 522 bp is shown. The four Greek key motifs are shown, including their N- and C-terminal amino acids and in relation to the coding exons. The A of the ATG initiation codon of the cDNA is counted as nucleotide 1; the first Met of the deduced amino acid sequence as amino acid 1. The Crygb gene of all mammals codes for an additional amino acid at the end of exon 2; the corresponding mRNA is 3 bp longer. The designation of the four Greek key motifs is according to Reference ³⁷ . (B) Alignments of the wild-type and mutated sequences: The mutations and the resultant changes in the amino acid sequence are shaded with white letters; the corresponding wild-type sequences are shaded with bold black letters. The altered restriction sites are underlined.

Figure 4.

 

Protein models of mutated γ-crystallins. Based on the crystal structure of the rat γE- and bovine γD-crystallin, 3-D models of the mutated γ-crystallins were calculated. The antiparallel β sheets are yellow and the α helices red; blue sections are looping regions. Amino acid substitutions are green. N, N terminus; C, C terminus. (A) The consensus γ-crystallin based on the rat γE- and the bovine γD-crystallin; the Greek key motifs are indicated by Roman numerals; (B) the novel point mutations (Arg43, Cryga ^ENU369 ; Pro45, Crygd ^ENU4011 , and Cryge ^ADD15306 ; Phe90, Crygd ^ENU910 ; Met126, Cryge ^ENU449 ); (C) C-terminal deletion in Crygc ^MNU8 ; (D) C-terminal deletion in Crygd ^K10 .

Figure 5.

 

Alterations in restriction sites associated with Cryg gene mutations and cataract formation. Genomic DNA from homozygous mutant or wild-type mice was amplified by PCR and digested by the corresponding restriction enzymes. Restriction maps of the PCR fragments are shown and the restriction sites indicated. The PCR fragments were analyzed by agarose gel electrophoresis with (+) or without (−) digestion by the indicated restriction enzyme (see the Results section for interpretation of gels). (A) Cryga fragment digested by HpaII for Cryga ^ENU369 mutation: The restriction site is present in the PCR-fragment of exons 1 and 2 in the four mutants only. (B) Crygc fragment digested by SduI for Crygc ^MNU8 mutation: The first SduI restriction site is missing in the 3′ part of Crygc, exon 3 of the mutants, resulting in two fragments of apparently the same size, which were not electrophoretically separated. (C) Crygd fragment digested by Bsp143I for the Crygd ^ENU910 mutation: The first Bsp143I restriction site in the PCR fragment containing exons 1 and 2 (Crygd) is missing only in the mutants, leading to two fragments of 226 and 237 bp, respectively. (D) Crygd fragment digested by ScaI for Crygd ^K10 mutation: The loss of the ScaI restriction site in exon 3 (Crygd) leads to an undigested fragment of 463 bp in all five mutants tested. (E) Cryge fragment digested by HphI for Cryge ^Z2 mutation: The 10-bp deletion results in loss of the second HphI restriction site in the PCR fragment containing exons 1 and 2 (Cryge), leading to two fragments of 198 and 356 bp. (F) Cryge fragment digested by Eco72 for the Cryge ^ENU449 mutation: The loss of the second Eco72 restriction site results in a fragment of 490 bp in the mutants. The slightly larger size of the C3H fragment compared with the other wild-type strains is due to a larger repeat in intron 1. M, marker; BL6, mouse strain C57BL/6; T, test stock mice; JF1, Japanese fancy mouse.

Table 2.

 

Polymorphisms in Mouse Cryg Genes

Gene	Reference Acc. No (Strain)	cDNA	Protein	Strain*
Cryga	NM_007774 (DBA/2)	201 A→G	No effect	T-stock; all mutant lines
Crygb	Z22573 (102/ElxC3H/El)F1	238 C→T	R80C	T-stock
		524 A→T	Y175F	T-stock
Crygc	Z22574 (102/ElxC3H/El)F1	111 A→C	R37R	T-stock; ENU4011
		384 G→A	E128E	T-stock
		513 A→C	V171V	ENU910
Crygd	NM_007776, †	95A→G	H32R	T-stock; all mutant lines
		154G→A; 156A→G	A52T	T-stock
		303/304 AG→GA	QV101/102QM	T-stock; all mutant lines
		321A→G	E107E
		488A→G	K163R	T-stock; all mutant lines
	AJ224342 (102× C3H)F1	Intron A: ΔGCCTT	No effect	JF1, C3H
Cryge	X57855 (101/ElxC3H/El)F1	139 C→G	Q47E	T-stock; all mutant lines
		154 A→G	T52A	T-stock
		189 T→C	Y63Y	T-stock; ENU4011
		Intron A: +11 A→G	No effect	129, ADD15306
		Intron B: reduced number of CTCAG repeats	No effect	T-stock
Crygf	Unpublished C3H	111 C→A	R37R	NM_027010 (B)
		234 T→C	S78S	NM_027010 (B)
		239 G→A	R80H	NM_027010 (B)
		468 C→T	D156D	T-stock; NM_027010 (B)
		484–486 GCC→TCA	A162S	T-stock, NM_027010 (B)
		501–503 GAG→TGA	LR167/168LE	T-stock, NM_027010 (B)

 A of the ATG start codon is counted as nt 1; Met encoded by the start codon is counted as amino acid 1.

*  B = C57BL/6J.

†  No specific strain is known (NIH general purpose mice).

Table 3.

 

List of Characterized Mutations at the Cryg Gene Cluster on Mouse Chromosome 1

Name	Previous Symbol	Phenotype (Graded Severity)	Paternal Treatment	Molecular Lesion	Consequence for Protein					Reference
											aa Alterations	pI	Greek Key Motif
Cryga ENU369	ENU369; Cat2 tol	Total cataract with vacuoles (severe)	ENU	Cryga: T127C	Trp43→Arg	pH 7.5–8.5	2nd not formed	This study
Cryga 1Neu	ENU-436	Diffuse nuclear opacity (medium)	ENU	Cryga: A222G	Asp74→Gly	pH 7.5–8.0	No effect	9
Crygb Nop	Nop	Nuclear opacity (medium)	Spontaneous	Crygb: Δ417–427/Ins4	Ser139: 6 new aa; stop	pH 7.6–5.5	4th not formed	9
Crygc Chl3	Chl3	Total and lamellar opacity (medium)	Chlorambucil	Crygc: Δ420–425	ΔG141, R142	pH 7.6–6.9	4th not formed	10
Crygc MNU8	MNU-8	Total opacity with vacuoles (medium)	MNU	Crygc: G471A	Trp157→stop 15 aa missing	pH 7.6–6.9	No effect	This study
Crygd ENU4011	ENU4011	Cloudy nuclear opacity (medium)	ENU	Crygd: T134C	Leu45→Pro	pH 6.6–7.0	2nd not formed	This study
Crygd Aey4	Aey4	Nuclear and cortical cataract (medium)	ENU	Crygd: T227A	Val76→Asp	pH 6.6–6.6	No effect	12
Crygd ENU910	ENU910	Diffuse total cataract (mild)	ENU	Crygd: A275T	Ile90→Phe	pH 6.6–7.0	No effect	This study
Crygd K10	K-10	Total cataract (medium)	Spontaneous	Crygd: C432G	Tyr144→stop 28 aa missing	pH 6.6–5.9	4th not formed	This study
Crygd Lop12	Lop12	Irregular nuclear lens opacity (medium)	Spontaneous	Crygd: G470A	Trp157→stop 18 aa missing	pH 6.6–6.2	No effect	11
Cryge Aey1	Aey1	Nuclear cataract (medium)	ENU	Cryge: A1T	new protein (13 kDa)	pH 7.7–9.5	No motif	15
Cryge ENU418	ENU418	Nuclear and lamellar cataract (medium)	ENU	Cryge: Intron A, A66G, lariat formation and splicing inhibited	Ile4: 153 new aa	pH 7.7–10.1	No motif	16
Cryge Z2	Z2	Total lamellar cataract (medium)	Spontaneous	Cryge: Δ12–21	Ile4: 119 new aa	pH 7.7–9.8	No motif	This study
Cryge Nz	Nzc; 116	Nuclear and zonular cataract (medium)	γ-ray	Cryge: Δ89T	Phe30: 96 new aa	pH 7.7–8.4	Motifs 2–4 not formed	14
Cryge ADD15306	ADD15306	Zonular cataract (medium)	ENU	Cryge: T134C;	Leu45→Pro	pH 7.7–7.1	2nd not formed	This study
Cryge ENU449	ENU449	Capsular opacity (mild)	ENU	Cryge: G376A	Val126→Met	pH 7.7–7.1	No effect	This study
Cryge Elo	Elo	Eye lens obsolescence (medium)	Spontaneous	Cryge: Δ403	Glu135: 11 new aa and stop	pH 7.7–6.4	4th not formed	13
Cryge t	R324; Cat2 t	Total opacity (severe)	X-ray	Cryge: C432G	Tyr144→stop	pH 7.7–6.4	4th not formed	9
Cryge Ns	K134; Scat	Suture cataract (mild)	Spontaneous	Cryge: Del >2kb	Hybrid protein?	?	4th motif affected?	50
Crygf Rop	Rop	Radial opacity (medium)	Procarbazine	Crygf: T113A	Val38→Glu	pH 7.1–6.7	4th not formed	17

 A of the ATG start codon is counted as nt 1; Met encoded by the start codon is counted as amino acid 1.

The authors thank Carmen Arnhold, Erika Bürkle, Sybille Frischholz, Bianca Hildebrand, Mareike Maurer, Brigitta May, and Irmgard Zaus for excellent technical assistance; Florian Giesert, Otmar Hainzl, Basile Siewe, and Susanne Wellnitz, who contributed to this study during their practical courses at the Technical University, Munich; and Utz Linzner, GSF-Institute of Experimental Genetics, for providing oligonucleotides.