Molecular Vision 2005; 11:569-581 <http://www.molvis.org/molvis/v11/a67/> Received 06 June 2005 | Accepted 18 July 2005 | Published 27 July 2005

Download Reprint

An ENU-induced mutation in Rs1h causes disruption of retinal structure and function

Monica M. Jablonski,^1,2 Claudia Dalke,³ XiaoFei Wang,^1,2 Lu Lu,^2,4 Kenneth F. Manly,^2,4 Walter Pretsch,⁵ Jack Favor,⁵ Machelle T. Pardue,^6,7 Eugene M. Rinchik,² Robert W. Williams,^2,4 Daniel Goldowitz,^2,4 Jochen Graw³
 
(The first two authors contributed equally to this publication)
 
 

¹Department of Ophthalmology and ⁴Center of Genomics and Bioinformatics, University of Tennessee Health Science Center, Memphis, TN; ²Tennessee Mouse Genome Consortium, Memphis, TN; ³GSF-National Research Center for Environment & Health, Institute of Developmental Genetics and ⁵Institute of Human Genetics, Neuherberg, Germany; ⁶Veterans Administration Medical Center, Atlanta, GA; ⁷Department of Ophthalmology, Emory University School of Medicine, Atlanta, GA

Correspondence to: Monica M. Jablonski, PhD, Department of Ophthalmology, Hamilton Eye Institute, University of Tennessee Health Science Center, 930 Madison, Suite 731, Memphis, TN, 38163; Phone: (901) 448-7572; FAX: (901) 448-1299; email: mjablonski@utmem.edu

Abstract

Purpose: The 44TNJ mutant mouse was generated by the Tennessee Mouse Genome Consortium (TMGC) using an ENU-based mutagenesis screen to produce recessive mutations that affect the eye and brain. Herein we present its retinal phenotype and genetic basis.

Methods: Fourth generation offspring (G₄) and confirmed mutants were examined using slit lamp biomicroscopy, funduscopy, histology, immunohistochemistry, and electroretinography (ERG). 44TNJ mutant mice were crossed to C3BLiA or DBA/2 mice for chromosomal mapping purposes. Linkage analysis by PCR-based microsatellite marker genotyping was used to identify the disease locus. The Rs1h cDNA and its genomic DNA were sequenced directly.

Results: The 44TNJ pedigree was the first mutant pedigree identified by the ocular phenotyping domain of the TMGC. Examination of the fundus revealed numerous small and homogeneous intraretinal microflecks in the peripapillary region, which became courser and more irregular in the periphery. Males were typically more affected than females. Histology and immunohistochemistry revealed a disruption of the lamination of the retina, particularly at both margins of the outer nuclear layer, along with reduced calbindin immunostaining. ERG analyses revealed reduced amplitudes of both a-waves and b-waves. Linkage analysis mapped the 44TNJ mutation to the X chromosome close to the marker DXMit117. Sequence analysis of the positional candidate gene Rs1h revealed a T->C exchange at the second base of intron 2 of the Rs1h gene.

Conclusions: We have generated and characterized a mutant mouse line that was produced using ENU-based mutagenesis. The 44TNJ pedigree manifests with photoreceptor dysfunction and concurrent structural and functional aberrations at the post-receptoral level. Genetic analysis revealed a mutation in Rs1h, making this the first murine model of X-linked retinoschisis in which the gene is expressed.

Introduction

X-linked retinoschisis (XLRS), the leading cause of juvenile macular degeneration affecting males, is caused by mutations in the RS1 gene [1]. While hemizygous affected males present with microcystic-like changes of the macula and schisis of inner retinal layers, peripheral lesions also are present in about 50% of cases [2-5]. Visual impairment is typically manifest by age 5 in boys. The progression of the disease is slow and complications such as vitreal hemorrhage, choroidal sclerosis, and retinal detachment can occur later in life. Female obligate carriers are typically asymptomatic with no clinical features, although Kaplan et al. [6] found that heterozygous females can have full expression of the retinal disease. The expression of XLRS is variable even within families. Moreover, the severity of disease expression is not mutation specific [7].

The human RS1 gene consists of six exons that encode for a 224 amino acid protein of 24 kDa. The Rs1 protein is soluble, secretory, and forms disulfide-linked dimers and octamers [8]. The predicted protein sequence contains three domains: a signal sequence encoded by exons 1 and 2 (amino acids 1-23); an Rs1 domain encoded by exon 3 (amino acids 23-62); and a discoidin domain encoded by exons 4-6 (amino acids 63-219). In addition, there is a five amino acid segment present on the C-terminal side of the discoidin domain [9]. To date, over 133 different mutations in RS1 have been documented [8] (see the RetinoschisisDB for details on specific mutations provided by the Retinoschisis Consortium). Most human mutations are in the discoidin domain with relatively few in signal or Rs1 domains, thereby suggesting that the discoidin domain and its putative role in cell adhesion are critical to the function of the protein [10-12].

Although the precise function of the protein remains incompletely understood, recent studies have determined that wild-type retinoschisin plays a role in both retinal cell adhesion [9,13,14] and synaptic integrity between photoreceptors and bipolar cells [13,15]. Recent elegant studies by several groups have provided evidence for three primary mechanisms that may be responsible for the loss of function of the retinoschisin protein, and therefore pathogenesis of XLRS [9]. The first of these involves misfolding of the discoidin domain, which negatively influences the putative adhesive properties of the protein [9]. The second mechanism is that the pathogenesis is caused by defective disulfide-linked subunit assembly of RS1 into dimers and octamers, which is also presumed to be critical for the proper function of the protein [8]. Lastly, an inability of RS1 to insert into endoplasmic reticulum membrane as part of the protein secretion process is a plausible mechanism for the retinal pathogenesis of XLRS [8,16]. Numerous studies [9,16,17] have demonstrated that in both COS-7 and Weri-Rb1 cultures, cells expressing constructs with mutations known to be pathogenic were retained intracellularly, rather than being secreted like the wild-type protein [16]. Thus these studies provide further mechanistic evidence for why the disease severity of XLRS is not mutation-dependent.

Within the last 20 years, the mouse has been established as an excellent model system for hereditary human diseases. The mouse has also a gene homologous to the human RS1, which is referred to as Rs1h (retinoschisis-1 homolog) [18]. Although two knockout (KO) models have been reported for the Rs1h gene [19,20], no spontaneous or ENU-induced mutant has been described up to now.

Within the TMGC, we have generated and characterized an ENU-induced mutant mouse pedigree with a retinal phenotype. Both retinal structure and function are affected in the 44TNJ mouse. Mutation analysis demonstrates a mutation in intron 2 of Rs1h, which leads to two novel splice variants. The 44TNJ pedigree has no other systemic or behavioral abnormalities, which is not surprising given that Rs1 is a retina-specific protein [19,20] (a complete listing of all phenotyping examinations performed on 44TNJ mice by the various domains of the TMGC is available at the TMGC website). This is the first description of an ENU-induced murine model of XLRS and will serve as a valuable resource to study the function of the Rs1h gene and its protein product.

Methods

Animal husbandry

Use of mice in this study was in compliance with the Guiding Principles in the Care and Use of Animals (DHEW Publication NIH 80-23), and was approved by the Animal Care and Use review board of the University of Tennessee Health Science Center and Oak Ridge National Laboratories (Oak Ridge, TN). The mice at the GSF-National Research Center for Environment and Health at Neuherberg, Germany were kept according to the Declaration of Helsinki and the regulations of the German Law on Animal Protection. Mice were maintained at 20 to 24 °C on a 14/10 h light/dark cycle at the University of Tennessee Health Science Center and on a 12/12 h light/dark cycle at GSF-National Research Center for Environment and Health. Animals were fed Agway Prolab 3000 or Altromin mouse chow and given tap water in glass bottles. While the majority of the mice examined in these studies were 10 weeks old, additional 44TNJ mice at various ages ranging from two weeks to 18 months were utilized in evaluating the ocular phenotype of this ENU-induced mutant mouse.

Generation of the ENU-induced mutant 44TNJ mice

Using an ENU mutagenesis strategy, the 44TNJ pedigree was derived from one C57BL/6JRn founder male mouse that was mated with a C3BLiA-Eh/+ female. Subsequently, a breeding scheme was followed such that genetically homozygous test class mice were generated by the third generation (G₃) of breeding. The fourth and subsequent generations should all carry the abnormal recessive mutation generated by the ENU-mutagenesis event. More details about this breeding protocol can be found at the neuromutagenesis domain of the TMGC website. Additional details regarding our mutagenesis program are described elsewhere [21-23]. 44TNJ mice are available to the scientific community through the TMGC website or through the Neuromice Consortium, a collaborative consortium that maintains and distributes the mutant mice generated through the TMGC, Northwestern University, and The Jackson Laboratories ENU-mutagenesis projects.

The 17TNK pedigree was used as a control throughout this manuscript and was found to be within normal limits in all tests by all domains of the TMGC. This pedigree was identical to the 44TNJ pedigree in all respects regarding background strain, mutagenesis with ENU, and husbandry.

Slit lamp and fundus examinations

Mice from the 44TNJ pedigree were clinically examined when they were approximately 10 weeks old. Mice were lightly anesthetized with an intraperitoneal injection of Avertin (1.25% 2,2,2-tribromoethanol and 0.8% tert-pentyl alcohol in water, 0.3 ml). The anterior segment was examined with a slit lamp biomicroscope (Carl Zeiss, Germany). 1% Cyclomydril ophthalmic drops (Alcon Pharmaceuticals, Fort Worth, TX) were used to dilate the pupil. The fundi were examined by indirect ophthalmoscopy and photographs were taken with a Kowa Genesis small animal fundus camera (Torrance, CA) with the aid of a 90 D condensing lens (Volk, Mentor, OH) as described previously [22,23]. 160T slide film (Kodak, Rochester, NY) was used for photodocumentation.

Fundus images from 112 10 week old 44TNJ mice were evaluated for ocular abnormalities. A grading system was used in which a normal fundus was graded as 1, a fundus that presented with small focal intraretinal microflecks throughout the posterior pole was graded as 2, and a fundus that presented with larger streak-like flecks in addition to intraretinal microflecks was graded as 3.

Light microscopy

Mice were sacrificed by cervical dislocation and eyes were removed immediately. All histological and immunohistochemical assessments were performed on the same eyes, rather than separate eyes being utilized for each type of analysis. Moreover, all images that are presented for a particular age of mouse are from the same eye and retinal area. The left eyes of eight 10 week old mice from the 44TNJ pedigree with fundus abnormalities were processed for paraffin embedding and histological/immunohistochemical evaluation. Both eyes from an 18 month old male mouse from the 44TNJ pedigree were similarly prepared. The left eyes from four 10 week old mice from the 17TNK pedigree were used as normal controls. Eyes were fixed in 4% paraformaldehyde in 0.06 M phosphate buffer for 24 h, then they were transferred to 0.06 M phosphate buffer. Embedding and sectioning were performed using standard protocols. The orientation of the eyes during the embedding process was controlled so that the anterior-posterior axis of the eye was parallel to the cutting surface of the block. Tissue sections taken from the posterior pole of each eye were stained with hematoxylin and eosin (H&E) using standard protocols. Sections were viewed on an Eclipse E800 microscope (Nikon Inc., Tokyo, Japan) equipped with a color camera (Photometrics, Tuscon AZ), and images of the retina within 1 to 2 mm of the optic nerve head were collected with MetaMorph software (Universal Imaging Corporation, West Chester, PA). Adobe Photoshop 7.0 (San Jose, CA) was used to prepare the final figures.

Immunohistochemistry

Paraffin sections were utilized for immunohistochemical analysis. After removal of the wax, sections were incubated in 1% H₂O₂ for 5 min and rinsed in phosphate buffered saline with 1% Tween-20 (PBS-T). Sections were incubated in primary antibody in PBS-T containing 5% goat serum overnight. The following primary antibodies were used in these studies: anti-GFAP (Immunon, 1:50 dilution) and anti-calbindin D-28K (1:250 dilution; Chemicon, Temecula, CA). These antibodies were selected due to the histological appearance of the retinas from 44TNJ mice; alterations in GFAP staining would be indicative of an overall nonspecific retinal pathology, while alterations in calbindin labeling would illustrate aberrations of the outer plexiform layer. Control sections were incubated in PBS-T containing 5% goat serum in the absence of primary antibody. Sections were then rinsed in PBS-T and incubated in anti-rabbit secondary antibody in PBS-T containing 5% goat serum for 1 h. The ABC kit (Vectorstain; Vector Laboratories, Inc., Burlingame, CA) was used to visualize the immunolabeling using the protocol suggested by the manufacturer. Sections were viewed on an Eclipse E800 microscope as described above.

Ultrastructure

The right eyes from three mice from the 44TNJ pedigree from three different ages were processed for more detailed evaluation of subcellular ultrastructure. The following ages of 44TNJ mice were utilized: 2 weeks, 14 weeks, and 38 weeks. These eyes were fixed in a mixed aldehyde consisting of 2% formalaldehyde and 2% glutaraldehyde in 0.06 M phosphate buffer for 24 h. The anterior segment of each eye was removed prior to further processing. Eyecups were dehydrated through a series of increasing ethanol concentrations and were embedding in Epon812 (EMS, Fort Washington, PA). Similar to eyes processed for paraffin sectioning, the orientation of the eyes was controlled so that the anterior-posterior axis of the eye was parallel to the cutting surface of the block. Thick sections were cut at 1 μm thickness through the posterior pole of the eye. Representative tissue sections were stained with Toluidine blue O. Sections were viewed on an Eclipse E800 microscope (Nikon Inc., Tokyo, Japan) as described above. Thin sections were cut and viewed on a JEOL JEM1200EX II electron microscope.

Electroretinography

Mice were dark-adapted overnight and anesthetized with a mixture of ketamine/xylazine (80 mg/kg and 16 mg/kg, respectively). Their pupils were dilated with 1% tropicamide, 1.5% cyclopentolate. A small wire loop electrode contacting the cornea through a layer of methylcellulose recorded the response from the retina, while a needle electrode placed in the cheek and tail served as reference and ground, respectively. Each mouse was placed in front of a Ganzfeld bowl (LKC Technologies, Inc., Gaithersburg, MD) that presented a series of flashes with increasing intensity (-3.0 to 2.1 log cd sec/m²). To isolate cone responses, an intensity series (-0.82 to 1.8 log cd sec/m²) was presented with a 30 cd/m² adapting field. Amplitude and implicit time measurements were recorded and compared between one eye of each 44TNJ mutant (n=4) and 17TNK control mouse (n=5). Statistical analysis of the relationship between amplitude and intensity was performed using repeated measure ANOVA with an α level of 0.05. To select for mutant mice, ERGs were performed at the GSF-National Research Center for Environment and Health, as described previously [24].

Gene mapping and mutation analysis

Gene mapping and mutation analysis were performed at both UTHSC and at the GSF-National Research Center for Environment and Health in Neuherberg. At UTHSC, F1 mice were generated by crossing 44TNJ affected female mice with C3BLiA male mice. F2 mice were produced by brother-sister matings. Mutant mice were identified by the fundus features and histological assessment. To narrow the genetic interval harboring the mutation, DNA was obtained from the tails of a set of 55 F2 mice. Microsatellite primer pairs were purchased from Invitrogen Life Technologies (Carlsbad, CA). PCR reactions were carried out in 96 well microtiter plates. A high-stringency touchdown protocol was used in which the annealing temperature was lowered progressively from 60 °C to 50 °C in 2 °C steps over the first 6 cycles [25]. After 30 cycles, products from the PCR reactions were run on 2.5% Metaphor agarose gels (FMC Bioproducts, Rockland ME), stained with ethidium bromide, and photographed. After an initial analysis using MIT microsatellite loci distributed across all autosomes and the X chromosome, the following microsatellite primer pairs were used to narrow the interval containing the mutation: DXMit89, DXMit166, DXMit144, DXMit1, DXMit61, DXMit172, DXMit173, DXMit38, and DXMit186. Similar to our previous study [26], genome-wide significance levels for assessing the confidence of the linkage statistics were estimated by comparing the peak likelihood ratio statistic (LRS) using the MapManager QTX program (version b17). The LRS can be converted to the conventional base-10 LOD score by dividing it by 4.61 (twice the natural logarithm of 10). The LRS is statistically convenient because its distribution is asymptotically a χ² distribution [27].

At the GSF-National Research Center for Environment and Health in Neuherberg, mice without the C3BLiA background were used to establish an independent line. Male carriers were mated to wild-type DBA/2 mice and F1 progeny were backcrossed to the 44TNJ mutant mice. Mutant mice were identified using ERG analysis. Genomic DNA was prepared from tail snips of 92 mice from the F2 progeny according to standard procedures. A linkage analysis was performed using the two microsatellite markers DXMit227 and DXMit117, located at 28.4 and 50.8 cM, respectively, using PCR primers produced at the GSF-National Research Center for Environment and Health. The cDNA was synthesized from mRNA isolated from entire eyes (stored at -80 °C) of adult mice according to standard procedures. For the PCR amplification of the mouse Rs1h gene, genomic or cDNA was used as a template, which was then amplified using the corresponding forward and reverse primers (Table 1). PCR products were sequenced commercially (SequiServe, Vaterstetten, Germany) after cloning into the pCRII Vector (Invitrogen, Karlsruhe, Germany) or after elution from the agarose gel using kits from Macherey & Nagel (Düren, Germany) and subsequent precipitation by ethanol and glycogen. To confirm the mutation, the PCR product (326 bp) from genomic DNA was digested by the restriction enzyme NlaIII

Results

Fundus phenotype

The fundi of control mice from the 17TNK pedigree were evenly pigmented (n=8; Figure 1A). In the fundi of male 44TNJ mice, there were many intraretinal microflecks throughout the posterior pole along with less-frequent, larger, more streak-like manifestations (Figure 1B). Most female 44TNJ mice presented with intraretinal microflecks, which were typically small and homogeneous within 3 disk diameters around the optic nerve head (Figure 1C). Application of our grading scale to the fundus photographs indicated that all male mice presented with an abnormal fundus phenotype, such that all male mice had a grade of either 2 or 3 (Figure 1D). Of the 45 male mice that were examined, 31% presented with only intraretinal flecks (grade 2), while the remaining 69% presented with both flecks and streaks (grade 3). Of the 67 female mice that were examined, 39% had a normal fundus appearance (grade 1), as compared to 45% that had only flecks (grade 2) and 16% that had both flecks and streaks (grade 3).

Retinal histology and immunohistochemical findings in the retinas of 10 week old male mice

The histological examination of retinas from 10 week old 17TNK male mice (n=4) were normal in all aspects. H&E stained tissue sections showed normal lamination, integrity of the RPE, and presence of photoreceptor outer segments (Figure 2A). In these same retinas, the immunoreaction product for GFAP was negligible in all layers of the 17TNK retinas (Figure 2B). Calbindin immunostaining revealed dense and continuous labeling in the outer plexiform layer with both the cell bodies and lateral processes of horizontal cells being labeled (Figure 2C).

In contrast, the histological examination of all eyes from eight 44TNJ male mice (Figure 2D) revealed a disruption of the lamination of the retina including migration of clusters of photoreceptor nuclei into the area of the inner and occasionally outer segments of the photoreceptors. These clusters of displaced nuclei very likely correspond to the microflecks that were evident on the fundus examination. The outer plexiform layer was periodically disrupted with cell bodies from the inner and outer nuclear layers abutting each other. Focal areas of inner nuclear layer splitting were also evident. At 10 weeks of age, the level of GFAP immunostaining was slightly elevated compared to control and radial patterns of Müller cell labeling were apparent (compare Figure 2E,B). Evaluation of calbindin-immunostained sections revealed a discontinuous and weak labeling pattern in the outer plexiform layer (Figure 2F).

Retinal histology and immunohistochemical findings in the retinas of 18 month old mice

Histological and immunohistochemical analyses of both eyes of one 18 month old 17TNK control mouse indicated that both the structure (Figure 2G-I), and the immunolabeling patterns of anti-GFAP (Figure 2J) and anti-calbindin (Figure 2K) were similar to 10 week old mice control mice. There was a slight increase in GFAP immunolabeling in the inner retina, which is a common phenomenon in the retinas of aged mice [22]. In both eyes of the aged 44TNJ mouse, however, there were some marked changes. There were many photoreceptor nuclei that migrated into the area of inner and outer segments, and the combined length of the inner/outer segments was very short compared to those of the age-matched 17TNK control mouse (Figure 2L,M, compare to Figure 2G,H). Moreover, in some areas, the outer plexiform layer had expanded deeper into the inner nuclear layer (Figure 2L,N). The areas of schisis within the inner nuclear layer were still prominent in the retinas from the aged mouse. There was a marked increase in the immunolabeling of Müller cells by the GFAP antibody (Figure 2O) compared to both the age-matched control (Figure 2J) and 10 week old 44TNJ mice (Figure 2B). The immunolabeling of horizontal cells with anti-calbindin antibodies demonstrated that the level of immunolabeling continued to be lower than its age-matched control (compare Figure 2P,K). In addition, it appeared that by 18 months, there was a migration of horizontal cells and their processes deeper into the inner nuclear layer, thus corresponding very well with the apparent expansion of the outer plexiform layer as shown histologically in Figure 2N.

Retinal ultrastructure at various ages

An evaluation of retinal structure was undertaken for the dual purpose of determining whether the aberrations in retinal lamination were due to a developmental defect and also if the retinal phenotype was progressive in nature. Our results indicated that as early as two weeks, the retina had all of the aberrant structural features that we documented at ten weeks of age (Figure 3A). Photoreceptor nuclei were present in the area of the inner segments, beyond the outer limiting membrane, and in the area of the outer plexiform layer. Moreover, there were areas of schisis present in the inner nuclear layer. At 14 weeks (Figure 3B), these same morphological characteristics were present in addition to rosette-like structures in the outer nuclear layer projecting deep into the inner nuclear layer. The small clusters of photoreceptor nuclei may correspond to the microflecks on the fundus examination, while the rosettes likely correspond to the streak-like patterns that are prominent on the fundus examination. At 38 weeks (Figure 3C), the morphological characteristics of the retina did not differ from that of retinas from younger mice, suggesting that up to 38 weeks, there is no marked histologic progression of the phenotype of 44TNJ mice.

Ultrastructural analysis of the retina of 14 week old male 44TNJ mice indicated that many aspects of retinal structure were compromised in this mutant mouse. The photoreceptor outer segments (Figure 3D) had focal gaps between adjacent outer segments with some membranous remnants filling those spaces. In areas where photoreceptor nuclei had migrated into the area of photoreceptor inner segments, there was a complete absence of the adherens junctions that are typically formed between Müller cells and the photoreceptors with which they are closely associated (Figure 3E). Moreover, the outer plexiform layer was disorganized and contained extracellular gaps filled with debris (Figure 3F).

ERG findings

Figure 4A shows the ERG intensity response series recorded under dark-adapted conditions from representative 17TNK control (left side) and 44TNJ mutant mice (right side) at four months of age. At the lowest flash intensity, the ERG response from both strains of mice contained the positive b-wave generated by the activity of the bipolar cells [28-31]. As the flash intensity increased, the negative a-wave generated by the photoreceptors [32,33] became visible and then both waves became larger and faster. Note the wavelets or oscillatory potentials on the leading edge of the b-wave in both strains. The only difference in the response between the two strains was a reduced amplitude of a- and b-waves in the 44TNJ mice. The difference in dark-adapted response amplitude can be seen more clearly in Figure 4C, which shows the dark-adapted a- and b-wave amplitudes across each flash intensity presented. While the 44TNJ mice showed near normal responses with dim flashes, there were significant reductions in amplitude at higher flash intensities (a-wave: F_6,36=9.48, p<0.0001; b-wave F_9,54=13.46, p<0.0001). Figure 4E demonstrates no difference in implicit time between responses from the 44TNJ mutant and 17TNK control mice.

Figure 4B shows representative the light-adapted responses, which isolated the cone response from a single 17TNK control and 44TNJ mutant mouse. The b-wave became clearly visible in the second step of the intensity series (-0.02 log cd sec/m²) and then grew in amplitude with increasing flash intensity. While the 44TNJ mice showed a slight reduction in light-adapted b-wave amplitude compared to 17TNK animals, the difference did not reach significance (Figure 4D; F_6,42=2.05, p=0.08). However, the implicit time for the light-adapted responses did show significantly faster responses for the 44TNJ mice, although this may be due to the variability in oscillatory potentials. (Figure 4F; F_6,42=5.394, p<0.001).

Genetic analysis

The DNA from 55 male affected F₂ mice were genotyped at UTHSC. The results indicated that the 44TNJ mutation mapped to the X Chromosome. The interval map was determined using two phenotypes, namely fundus appearance and retinal structure derived from H&E stained paraffin sections. This locus includes the markers from DXMit61 (30.9 cM) to DXMit186 (69.0 cM); the corresponding LRS scores were 11.4 and 24.0, respectively. A maximum LRS score of 24 was obtained for linkage to DXMit186 (χ² for males=40).

For fine mapping of the X-chromosome, 92 mice from the F₂ progeny were genotyped at the GSF using the two markers DXMit227 and DXMit117, located at 28.4 and 50.8 cM, respectively. Highly significant linkage was found to marker DXMit117 (χ² for males=23.5). Calculating the recombination frequency, it was estimated that the mutation should be located 21±6 cM distal of this marker, since the proximal marker DXMit227 at the position 28.4 cM did not show a significant linkage (χ² for males=1.4).

Because of the critical position of about 70 cM, the retinoschisis-1 homolog gene Rs1h was tested as candidate of the 44TNJ mutation. Amplification of the Rs1h cDNA from 44TNJ mutants revealed two PCR products instead of a single one, as was found in the wild-type mouse (Figure 5A). Sequencing of the larger PCR product of 44TNJ mice revealed a "mixed" sequence (Figure 5B). One part was the wild-type sequence, the other part showed a frameshift caused by a 10 bp insertion. This 10 bp insertion was identical to the first base of intron 2 with the exception of a T to C change at the second base. The sequence analysis of the smaller fragment of the 44TNJ-Rs1h cDNA revealed a 26 bp deletion also at the transition from exon 2 to intron 2 (Figure 5C).

To confirm this mutation at the genomic level, we amplified a 326 bp fragment of genomic DNA covering this critical position. The T to C exchange creates a new restriction site for NlaIII, which was not present in the wild-type (Figure 6). We confirmed the presence of this novel restriction site in five male hemizygous mice with the mutation, however, we did not find this restriction site in any of the five different wild-type controls of C57BL/6J, C3H, BALB/c, 129, and JF1 mice (Figure 6). Therefore, the 44TNJ mutation should be referred to as a new allele of the Rs1h gene, Rs1h^44TNJ.

Computer assisted translation of the novel Rs1h splice products in the 44TNJ mutants revealed the formation of premature stop codons because of the frameshift caused by both the 10 bp insertion and the 26 bp deletion (Figure 7). In case of the 26 bp deletion (corresponding to the entire exon 2), the new open reading frame reaches a new stop codon immediately at the beginning of exon 3, thus resulting in a truncated protein that consists only of the first 17 amino acids plus an additional new one. In case of the 10 bp insertion after exon 2, the first 26 amino acids were present, followed by four new amino acids and a premature stop codon at the beginning of exon 3. In both cases, after only a few bases, other open reading frames were present, which might be used also by the translation machinery. Further experiments are needed to clarify whether the short mutated N-terminal Rs1h peptides are present in the retina or whether the corresponding mutant mRNAs are degraded because of nonsense-mediated decay. Similarly, future experiments will address the question of whether the proteins encoded by the new open reading frames are present in the retina.

Discussion

Herein we characterize and present the ocular phenotype of an ENU-induced mutant mouse with a novel mutation located in an intronic sequence of Rs1h. While there have been previous reports of Rs1h KO mice [19,20], this is the first presentation of a murine model with a mutation in Rs1h. The previous Rs1h KO mice presented with the morphologic phenotypes of retinal layer disruption indicated by displaced photoreceptor nuclei and dissection or schisis of the inner nuclear layer of the retina. Moreover, these mice had a disruption of synaptic structure in the outer plexiform layer where photoreceptors synapse on second order neurons, such as bipolar cells. As expected, given the morphologic phenotype, the b-wave of the scotopic ERG response was significantly reduced in the Rs1h KO mice generated by both groups [19,20]. Zeng et al., however, also demonstrated a reduced a-wave amplitude in their KO mice, which correlated well with the shortened photoreceptor inner and outer segments that they documented [19,20].

The morphological and functional phenotype of 44TNJ mice is strikingly similar to both Rs1h KO mice. The 44TNJ mice have similar displacement of photoreceptor nuclei both distally and proximally. Moreover, we demonstrate a reduced and discontinuous calbindin labeling pattern that correlates well with the disruption of the outer plexiform layer as seen morphologically. Coupled with this are dissections or splitting of the inner nuclear layer. ERG analyses indicate a reduced contribution of both photoreceptors and second order neuron function in these mutant mice.

Although not presented in the previous descriptions of Rs1h KO mice, we demonstrate the presence of intraretinal flecks upon fundus examination in all males, and in 61% of the females. Often the severity of the fundus features was increased by the presence of larger intraretinal streaks. Morphologically, the flecks and streaks likely correlate with the undulating pattern of the outer nuclear layer and the presence of rosettes. The phenotype of the 44TNJ mice parallels nicely the disease expression variations that have been documented in human families [7]. Female obligate carriers of mutations in RS1 are typically asymptomatic. Our fundus examinations of 67 female test class mice from the 44TNJ pedigree indicate that 61% of these mice have fundus features similar to that of the male mice. Although this cohort of mice were not genotyped for mutations in Rs1h, these females were likely homozygous for mutations in Rs1h, which is due to the breeding scheme that was utilized to generate the G₄ test class mice.

Similar to the human condition [3,4], male hemizygous mutant mice have a retinal phenotype at an early age. The morphological aberrations of the retina are present as early as two weeks of age in the 44TNJ mice and do not appear to progress up to 38 weeks. At 18 months, however, there appears to be a worsening of the morphological phenotype along with an elevated GFAP response of the Müller glia, which is not due to aging alone [22]. Because the results from the aged mice were obtained from only one mouse of each pedigree, they should not be considered definitive until confirmed in other mice of a similar age. Nonetheless, this result parallels the progression of the disease in humans in which retinal detachments and a further decrease in vision can occur later in life [3].

The mutation in 44TNJ mouse is a point mutation in Rs1h with a T->C substitution in intron 2 at position 2. The net effect of this mutation is introduction of an alternative splice site and the expression of two new splice products in the retinas of the 44TNJ mutant. The first alternative transcript has a 10 bp insertion between positions 88 and 89, whereas second alternative transcript has a 26 bp deletion at position 62. While neither splice variant encodes for a stop codon at the position of the mutation, both lead to a frameshift and to a premature stop codon slightly downstream of the mutation. Because these stop codons are introduced very early in the sequence of Rs1h, it is predicted that only very short peptides will be produced by the alternative splice sites. The predicted truncated protein products of the mutation likely explain the similarity of the phenotype of the 44TNJ mouse with a point mutation of Rs1h and the Rs1h KO models previously described [19,20]. The putative abbreviated peptides present in the retinas of 44TNJ mice likely undergo nonsense-mediated decay of the mutated mRNAs [34], thus equating the 44TNJ mouse to a hypomorph. Another, although less likely, explanation may be that the gene product of the intronic mutation exerts a "dominant-negative" effect. These hypotheses will need to be experimentally tested before any conclusions can be derived as to the mechanism(s) by which a mutation in intron 2 of Rs1h leads to disruption of retinal structure and function.

In humans, approximately 80% of the mutations are missense mutations in the discoidin domain [12,35], which is encoded by exons 4-6. The other 20% of the mutations are present in exons 1-3 and result in altered transcription of the gene with premature truncation or loss of Rs1 expression [36]. While the vast majority of mutations in XLRS are located in exons, several recent reports have documented mutations in intronic sequences in the human [37-41]. The 44TNJ mouse with a point mutation in intron 2 is a novel model for XLRS and would allow for the study of the function of the Rs1 protein that is completely lacking the discoidin domain. Moreover, because the mutation in the 44TNJ mouse allows for truncated Rs1 protein to be expressed, the disease mechanisms underlying the retinal pathology should be very similar to that of the human disease, thus allowing for many of the recently proposed pathological mechanisms that stem from in vitro studies to be confirmed in vivo and studied in greater depth.

Acknowledgements

We thank Mareike Maurer and Kathy Troughton for excellent technical assistance. The work was supported in part by U01-MH61971 (to DG and MMJ), the German National Genome Network (NGFN; 01GR0103 to JG), the Department of Veterans Affairs (to MTP), and unrestricted grants from Research to Prevent Blindness (New York, NY) to the Department of Ophthalmology at the University of Tennessee Health Science Center (Memphis, TN) and the Department of Ophthalmology, Emory University School of Medicine (Atlanta, GA).

References

1. Sauer CG, Gehrig A, Warneke-Wittstock R, Marquardt A, Ewing CC, Gibson A, Lorenz B, Jurklies B, Weber BH. Positional cloning of the gene associated with X-linked juvenile retinoschisis. Nat Genet 1997; 17:164-70.

2. Kellner U, Brummer S, Foerster MH, Wessing A. X-linked congenital retinoschisis. Graefes Arch Clin Exp Ophthalmol 1990; 228:432-7.

3. George ND, Yates JR, Moore AT. X linked retinoschisis. Br J Ophthalmol 1995; 79:697-702.

4. George ND, Yates JR, Moore AT. Clinical features in affected males with X-linked retinoschisis. Arch Ophthalmol 1996; 114:274-80.

5. Sieving PA. Juvenile retinoschisis. In: Traboulsi E, editor. Genetic diseases of the eye. New York: Oxford University Press; 1998. pp. 347-55.

6. Kaplan J, Pelet A, Hentati H, Jeanpierre M, Briard ML, Journel H, Munnich A, Dufier JL. Contribution to carrier detection and genetic counselling in X linked retinoschisis. J Med Genet 1991; 28:383-8.

7. Eksandh LC, Ponjavic V, Ayyagari R, Bingham EL, Hiriyanna KT, Andreasson S, Ehinger B, Sieving PA. Phenotypic expression of juvenile X-linked retinoschisis in Swedish families with different mutations in the XLRS1 gene. Arch Ophthalmol 2000; 118:1098-104.

8. Wu WW, Wong JP, Kast J, Molday RS. RS1, a discoidin domain-containing retinal cell adhesion protein associated with X-linked retinoschisis, exists as a novel disulfide-linked octamer. J Biol Chem 2005; 280:10721-30.

9. Wu WW, Molday RS. Defective discoidin domain structure, subunit assembly, and endoplasmic reticulum processing of retinoschisin are primary mechanisms responsible for X-linked retinoschisis. J Biol Chem 2003; 278:28139-46.

10. Baumgartner S, Hofmann K, Chiquet-Ehrismann R, Bucher P. The discoidin domain family revisited: new members from prokaryotes and a homology-based fold prediction. Protein Sci 1998; 7:1626-31.

11. Vogel W. Discoidin domain receptors: structural relations and functional implications. FASEB J 1999; 13:S77-82.

12. Functional implications of the spectrum of mutations found in 234 cases with X-linked juvenile retinoschisis. The Retinoschisis Consortium. Hum Mol Genet 1998; 7:1185-92.

13. Molday LL, Hicks D, Sauer CG, Weber BH, Molday RS. Expression of X-linked retinoschisis protein RS1 in photoreceptor and bipolar cells. Invest Ophthalmol Vis Sci 2001; 42:816-25.

14. Reid SN, Yamashita C, Farber DB. Retinoschisin, a photoreceptor-secreted protein, and its interaction with bipolar and muller cells. J Neurosci 2003; 23:6030-40.

15. Takada Y, Fariss RN, Tanikawa A, Zeng Y, Carper D, Bush R, Sieving PA. A retinal neuronal developmental wave of retinoschisin expression begins in ganglion cells during layer formation. Invest Ophthalmol Vis Sci 2004; 45:3302-12.

16. Wang T, Waters CT, Rothman AM, Jakins TJ, Romisch K, Trump D. Intracellular retention of mutant retinoschisin is the pathological mechanism underlying X-linked retinoschisis. Hum Mol Genet 2002; 11:3097-105.

17. Grayson C, Reid SN, Ellis JA, Rutherford A, Sowden JC, Yates JR, Farber DB, Trump D. Retinoschisin, the X-linked retinoschisis protein, is a secreted photoreceptor protein, and is expressed and released by Weri-Rb1 cells. Hum Mol Genet 2000; 9:1873-9.

18. Gehrig AE, Warneke-Wittstock R, Sauer CG, Weber BH. Isolation and characterization of the murine X-linked juvenile retinoschisis (Rs1h) gene. Mamm Genome 1999; 10:303-7.

19. Weber BH, Schrewe H, Molday LL, Gehrig A, White KL, Seeliger MW, Jaissle GB, Friedburg C, Tamm E, Molday RS. Inactivation of the murine X-linked juvenile retinoschisis gene, Rs1h, suggests a role of retinoschisin in retinal cell layer organization and synaptic structure. Proc Natl Acad Sci U S A 2002; 99:6222-7.

20. Zeng Y, Takada Y, Kjellstrom S, Hiriyanna K, Tanikawa A, Wawrousek E, Smaoui N, Caruso R, Bush RA, Sieving PA. RS-1 gene delivery to an adult Rs1h knockout mouse model restores ERG b-Wave with reversal of the electronegative waveform of X-linked retinoschisis. Invest Ophthalmol Vis Sci 2004; 45:3279-85.

21. Goldowitz D, Frankel WN, Takahashi JS, Holtz-Vitaterna M, Bult C, Kibbe WA, Snoddy J, Li Y, Pretel S, Yates J, Swanson DJ. Large-scale mutagenesis of the mouse to understand the genetic bases of nervous system structure and function. Brain Res Mol Brain Res 2004; 132:105-15.

22. Johnson DK, Rinchik EM, Moustaid-Moussa N, Miller DR, Williams RW, Michaud EJ, Jablonski MM, Elberger A, Hamre K, Smeyne R, Chesler E, Goldowitz, D. Phenotype screening for genetically determined age-onset disorders and increased longevity in ENU-mutagenized mice. Age 2005; 27:75-90.

23. Jablonski MM, Wang XF, Lu L, Miller DR, Rinchik EM, Williams RW, Goldowitz D. The Tennessee Mouse Genome Consortium: Identification of ocular mutants. Vis Neurosci. In press 2005.

24. Dalke C, Loster J, Fuchs H, Gailus-Durner V, Soewarto D, Favor J, Neuhauser-Klaus A, Pretsch W, Gekeler F, Shinoda K, Zrenner E, Meitinger T, Hrabe de Angelis M, Graw J. Electroretinography as a screening method for mutations causing retinal dysfunction in mice. Invest Ophthalmol Vis Sci 2004; 45:601-9.

25. Don RH, Cox PT, Wainwright BJ, Baker K, Mattick JS. 'Touchdown' PCR to circumvent spurious priming during gene amplification. Nucleic Acids Res 1991; 19:4008.

26. Jablonski MM, Lu L, Wang X, Chesler EJ, Carps E, Qi S, Gu J, Williams RW. The ldis1 lens mutation in RIIIS/J mice maps to chromosome 8 near cadherin 1. Mol Vis 2004; 10:577-87 <http://www.molvis.org/molvis/v10/a70/>.

27. Knott SA, Haley CS. Maximum likelihood mapping of quantitative trait loci using full-sib families. Genetics 1992; 132:1211-22.

28. Stockton RA, Slaughter MM. B-wave of the electroretinogram. A reflection of ON bipolar cell activity. J Gen Physiol 1989; 93:101-22.

29. Xu X, Karwoski CJ. Current source density analysis of retinal field potentials. II. Pharmacological analysis of the b-wave and M-wave. J Neurophysiol 1994; 72:96-105.

30. Robson JG, Frishman LJ. Response linearity and kinetics of the cat retina: the bipolar cell component of the dark-adapted electroretinogram. Vis Neurosci 1995; 12:837-50.

31. Kofuji P, Ceelen P, Zahs KR, Surbeck LW, Lester HA, Newman EA. Genetic inactivation of an inwardly rectifying potassium channel (Kir4.1 subunit) in mice: phenotypic impact in retina. J Neurosci 2000; 20:5733-40.

32. Penn RD, Hagins WA. Signal transmission along retinal rods and the origin of the electroretinographic a-wave. Nature 1969; 223:201-4.

33. Robson JG, Frishman LJ. Dissecting the dark-adapted electroretinogram. Doc Ophthalmol 1998-99; 95:187-215.

34. Holbrook JA, Neu-Yilik G, Hentze MW, Kulozik AE. Nonsense-mediated decay approaches the clinic. Nat Genet 2004; 36:801-8.

35. Tantri A, Vrabec TR, Cu-Unjieng A, Frost A, Annesley WH Jr, Donoso LA. X-linked retinoschisis: a clinical and molecular genetic review. Surv Ophthalmol 2004; 49:214-30.

36. Hiriyanna K, Singh-Parikshak R, Bingham EL, Sieving PA. Searching for genotype-phenotype correlations in X-linked juvenile retinoschisis. In: Anderson RE, LaVail MM, Hollyfield JG, editors. New insights into retinal degenerative diseases. Proceedings of the 9th International Symposium on Retinal Degeneration; 2000 Oct 9-14; Durango, CO. New York: Kluwer Academic/Plenum Publishers; 2001. p. 45-53.

37. Hiriyanna KT, Bingham EL, Yashar BM, Ayyagari R, Fishman G, Small KW, Weinberg DV, Weleber RG, Lewis RA, Andreasson S, Richards JE, Sieving PA. Novel mutations in XLRS1 causing retinoschisis, including first evidence of putative leader sequence change. Hum Mutat 1999; 14:423-7.

38. Mashima Y, Shinoda K, Ishida S, Ozawa Y, Kudoh J, Iwata T, Oguchi Y, Shimizu N. Identification of four novel mutations of the XLRS1 gene in Japanese patients with X-linked juvenile retinoschisis. Mutation in brief no. 234. Online. Hum Mutat 1999; 13:338.

39. Shinoda K, Ishida S, Oguchi Y, Mashima Y. Clinical characteristics of 14 japanese patients with X-linked juvenile retinoschisis associated with XLRS1 mutation. Ophthalmic Genet 2000; 21:171-80.

40. Stanga PE, Chong NH, Reck AC, Hardcastle AJ, Holder GE. Optical coherence tomography and electrophysiology in X-linked juvenile retinoschisis associated with a novel mutation in the XLRS1 gene. Retina 2001; 21:78-80.

41. Sato M, Oshika T, Kaji Y, Nose H. Three novel mutations in the X-linked juvenile retinoschisis (XLRS1) gene in 6 Japanese patients, 1 of whom had Turner's syndrome. Ophthalmic Res 2003; 35:295-300.