A New Fgf10 Mutation in the Mouse Leads to Atrophy of the Harderian Gland and Slit-Eye Phenotype in Heterozygotes: A Novel Model for Dry-Eye Disease?

Abstract

purpose. The purpose of the present study was to characterize a new slit-eye phenotype in the mouse.

methods. Genomewide linkage analysis was performed, and a candidate gene was sequenced. Eyes of the mutants were described morphologically, histologically, and by in situ hybridization. To allow morphologic and functional studies of the retina, mutants were outcrossed to C57BL/6.

results. Within an ongoing ethyl-nitrosourea mutagenesis screen with C3HeB/FeJ mice, the authors identified a new mutant (referred to as Aey17) showing a slit-eye phenotype in heterozygotes; homozygous mutants are not viable because of major developmental defects. This mutation was mapped to the distal end of mouse chromosome 13, suggesting Fgf10 (encoding the fibroblast growth factor 10) as a candidate gene. An A→G transition in the penultimate base of the first intron of Fgf10 leading to aberrant splicing with an additional 49 bp in exon 2 and to a frameshift with a premature stop codon after 54 new amino acids was identified. Histologic analysis of the major ocular tissues (cornea, lens, retina) did not reveal major alterations compared with the wild type, but the size of the Harderian gland was remarkably reduced in heterozygotes. Although Fgf10 was expressed in the developing retina, neither electroretinography nor the virtual drum indicated any abnormalities in heterozygous mutants; overall eye size was identical in wild types and heterozygotes.

conclusions. The mutation in the Fgf10 gene leads to a dominant slit-eye phenotype caused by atrophy of the Harderian gland.

The small eye phenotype in the mouse is rather frequent and reflects a genetically diverse disorder caused by major defects in early eye development (for a review, see Graw ¹ ). The major players are Mitf (encoding the microphthalmia-associated transcription factor ^{2 3 4 5} ) and Pax6 (paired-box gene 6 ^{6 7 8 9} ). In addition to these major players characterized by several affected alleles, Pitx3, Maf, and Sox2 have been shown to lead to similar phenotypes. However, only a few alleles of these genes have been characterized until now (for a recent review, see Ref. ¹⁰ ). Recently, we reported a new small-eye phenotype in the mouse caused by a mutation in the gene Gjf1 coding for a connexin-like protein. ¹¹  

In the mouse, the small-eye phenotype might be misclassified because of a rare alteration caused by changes in the eyelids leading to a slit-eye phenotype but eyes of regular size. Such a phenotype has been described in the mouse in two mutants, a PPP2CA transgenic mouse ¹² and the Muc1 knockout mouse. ¹³ In both cases, the slit-eye phenotype is associated with Harderian gland aplasia or malfunction, resulting frequently in recurrent inflammation of ocular tissues. A similar sequence of events has been observed in the dry-eye syndrome (keratoconjunctivitis sicca), which is considered a complex multifactorial disease. ¹⁴  

Here we present a novel dominant slit-eye mouse mutant that was identified in an ethyl-nitrosourea (ENU) mutagenesis screen with C3H mice ¹⁵ ; the mutation is referred to as Aey17. Similar to some of the mutants described, Aey17 homozygotes are not viable. However, genetic analysis identified a new gene responsible for this phenotype, Fgf10 (encoding the fibroblast factor 10), and added a new gene to the still short list of slit-eye causing mutations. It is the first point mutation in the Fgf10 gene of the mouse and is offered to the community as a new model for the dry-eye syndrome, particularly for the dominant aplasia of lacrimal and salivary glands (ALSG). 

Materials and Methods

Materials

Chemicals were obtained from Merck (Darmstadt, Germany) or Sigma Chemical (Deisenhofen, Germany). The enzymes used for cloning and reverse transcription were from Roche (Mannheim, Germany). 

Mice

Mice were kept under specific pathogen-free conditions at the Helmholtz Center Munich and were monitored within the ENU mouse mutagenesis project. ^{15 16} Animals were used in accordance with the German Law on Animal Protection and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Male C3HeB/FeJ mice were treated with ENU (3 × 100 mg/kg body weight) at the age of 10 weeks in accordance with Ehling et al. ¹⁷ and were mated to untreated female C3HeB/FeJ mice. Offspring of the ENU-treated mice were screened at the age of 3 weeks for abnormalities of the eye (Aey). Any mouse with a phenotypic deviation was tested for a dominant mode of inheritance. 

For pathologic investigation, mice were killed with CO₂, analyzed macroscopically, and weighed. The thymus and left lobe of the liver were measured. The Harderian gland was identified, isolated, and submitted separately for subsequent histologic examination. All organs were fixed in 4% buffered formalin and embedded in paraffin for histologic examination. Sections (2-μm thick) from skin, heart, muscle, lung, brain, cerebellum, thymus, spleen, cervical lymph nodes, thyroid, parathyroid, adrenal gland, stomach, intestine, liver, pancreas, kidney, reproductive organs, urinary bladder, and Harderian gland were cut and stained with hematoxylin and eosin (H&E). Additionally, periodic acid Schiff (PAS) reaction and Masson trichrome staining, as well as immunostaining for a pan-cytokeratin (CK) marker (clone AE1/AE3; DCS Innovative Diagnostik-System, Hamburg, Germany), were used for better visualization of the Harderian gland. Immunohistochemistry was performed using an automated immunostainer with a DAB detection kit (iVIEW; Ventana Medical System, Inc., Tucson, AZ) according to the company’s protocols for open procedures with slight modifications. 

Mapping

Heterozygous carriers (first generation) were mated to wild-type C57BL/6J mice, and the offspring (second generation) were backcrossed to wild-type C57BL/6J mice. DNA was prepared from tail tips of affected offspring of the third generation (G3) according to standard procedures. DNA was adjusted to a concentration of 50 ng/μL. For genomewide linkage analysis of the Aey17 mutation, several microsatellite markers were used for each autosome. ¹⁸  

PCR and Sequencing

For the molecular analysis, RNA from the heads of Aey17 ^−/− or wild-type embryos (C3HeB/FeJ) was isolated at day 12.5 of embryonic development. RNA samples were reverse transcribed to cDNA (Ready-to-Go kit; Pharmacia Biotech, Freiburg, Germany); genomic DNA was isolated from tail tips or spleens of wild-type C3HeB/FeJ and C57BL/6J mice or homozygous mutants according to standard procedures. For RT-PCR amplification, primers were selected from the EMBL-GenBank databases (Table 1) . 

PCR was performed using a thermocycler (Clontech, Heidelberg, Germany; PE-Biosystems, Weiterstadt, Germany). PCR products were analyzed on a 1% agarose gel. Sequencing was performed commercially (SequiServe, Vaterstetten, Germany; GTAC, Konstanz, Germany) or in the Genome Analysis Center of the Helmholtz Center Munich-German Research Center for Environmental Health, either after direct purification of the PCR products (Nucleospin Extract II; Machery-Nagel, Düren, Germany) or after isolation of the DNA from the gel using an extraction kit (Qia-Quick; Qiagen, Hilden, Germany) or after cloning into the pCR-2.1 cloning vector (Invitrogen, Karlsruhe, Germany). 

Phenotypic Characterization of Aey17 Mutants

Eyes of the mutants were examined with a slit lamp when they were 3 weeks of age ¹⁹ and by funduscopy, electroretinography, or optical laser interferometry when they were 10 to 14 weeks of age, as described previously. ^{20 21 22}  

Vision tests were performed between 9 am and 4 pm using a virtual optomotor system (Cerebral Mechanics, Lethbridge, AB, Canada), as described previously. ²³ Briefly, a rotating cylinder covered with a vertical sine wave grating was calculated and drawn in virtual three-dimensional space on four computer monitors facing to form a square. Visually unimpaired mice track the grating with reflexive head and neck movements (head-tracking) while standing unrestrained on a platform in the center of the square. Visual acuity of the tested mouse was quantified by a stepwise increase of the spatial frequency (0.3, 0.319, 0.342, 0.364, 0.386, 0.414, 0.439, 0.469, and 0.5 cyc/deg) until an optomotor response could no longer be detected. Highest spatial frequency that the tested mouse could track was identified as the threshold. To avoid habituation, rotation direction was changed several times during measurement. Rotation speed and contrast were set to 12 d/s and 100%, respectively. Thresholds of C57BL/6J wild-types and Aey17 heterozygotes were compared using the Mann-Whitney U test. 

Eyes at embryonic day 12.5 (E12.5), E15.5, or E17.5 after fertilization and at postnatal day 1 (P1) or P7 were analyzed histologically. Embryos, their heads, or their entire eyes (at P1 and P7) were fixed for 3 hours in Carnoy solution and embedded in plastic medium (JB-4; Polysciences Inc., Eppelheim, Germany) according to the manufacturer’s protocol. Sectioning was performed with an ultramicrotome (OMU3; Reichert-Jung, Walldorf, Germany). Serial transverse 3-μm sections were cut with a glass knife and stained with methylene blue and basic fuchsin. Sections were evaluated with a light microscope (Axioplan-2; Zeiss, Munich, Germany). Images were acquired by means of a scanning camera (AxioCam; Zeiss) and were imported into an image-processing program (Photoshop 10.0; Adobe, Unterschleissheim, Germany). 

In Situ Hybridization

In situ hybridization of sections from different embryonic stages were performed essentially as described. ²⁴ Briefly, the embryos were fixed in paraformaldehyde and embedded in Jung Histowax (Cambridge Instruments, Nussloch, Germany). Sections (7–10 μm) were cut with a microtome (RM-2065; Leica, Nussloch, Germany) and mounted onto slides. Hybridization was performed with an Fgf10-specific probe amplified by PCR and cloned into a vector (pCRII-TOPO; Invitrogen, Karlsruhe, Germany). Primers for PCR-amplification were Fgf10-L1 and Fgf10-R1 (Table 1) . Sections were evaluated with a light microscope (Axioplan; Zeiss), and images were acquired as described. 

Results

Phenotype

The Aey17 mutant was detected during an ENU mutagenesis screen because of its dominant slit-eye phenotype (Fig. 1a) ; subsequently, it was established as a heterozygous line. Heterozygotes were characterized by severely reduced Harderian glands (Fig. 1b) . Homozygous mutants are not viable; some offspring were born without arms and legs; an example of a mature embryo is shown in Figure 1c . 

Because of the slit-eye phenotype (Fig. 1a) , we suggested microphthalmia and analyzed eye size using the optic laser interference system. Surprisingly, the size of the entire eye was not significantly different between the wild-type C3H mice (3.516 ± 0.031 mm; n = 18) and the heterozygous (3.511 ± 0.023 mm; n = 11) mutants. Similarly, the cornea, lens, and vitreous were not significantly different between wild-type and mutant mice (data not shown); there was also no difference between males and females. The slit-eye phenotype was confirmed without any contribution of microphthalmia. 

The Harderian gland displayed in the wild-type animals the characteristic histologic aspect of a branched tubuloalveolar gland (Fig. 2a) . The alveoli were lined by a single layer of columnar cells with round basal nuclei and dark eosinophilic cytoplasm. Alveolar lumens sometimes contained dark porphyrin pigment concretions. Dark pigment was also seen in the thin connective tissue septa between the alveoli. All cells displayed a diffuse positive reaction to CK AE1/AE3 (Fig. 2e , inset). In the mutant animals, however, there was a subtotal atrophy of the gland that was represented by a fibrotic, pigmented mass, with only a few remaining alveoli lined by epithelial cells with cytoplasmic fat vacuoles (foamy cells; Figs. 2b 2c 2d ). Residual immunoreactivity for CK AE1/AE3 was observed in the foamy cells (Fig. 2e) , whereas the fibrotic mass was completely devoid of any CK-positive cell (Fig. 2f) . All other analyzed organs did not show any pathologic alteration (not shown). 

Linkage Analysis and Molecular Characterization

The mutation arose in an ENU-treated C3HeB/FeJ mouse; after outcross/backcross to C57BL/6J mice, we tested in a genomewide linkage analysis 87 F₂ mice with the Aey17 phenotype. We mapped the mutation to the distal end of mouse chromosome 13 between the markers D13Mit77 (at position 117.7 MB) and D13Mit35 (at position 120.1 MB); the haplotype analysis is shown in Figure 3a . The physical map of this region (Fig. 3b)suggested Fgf10 as a strong candidate gene (at position 119.5 MB). Sequence analysis of the Fgf10 transcript revealed an additional 49 bp at the beginning of exon 2. The causative mutation is an A→G transition in the penultimate base of the first intron (g.IVS1–2A>G; Fig. 3c ). Because it destroys a canonical AG/GT splice site, a similar splice site 49 bp upstream is used in the mutants (AG/CT), leading to the observed insertion in the mature mRNA (Fig 3d) . The mutation in the Aey17 mice does not destroy an existing restriction site and does not create a new one; therefore, the corresponding site was sequenced in 10 different inbred strains of mice (102/El, AKR, C3H, C57BL/6, CBA, CFW, DBA, JF1, NMRI, SEC). Given that the mutation was present in all heterozygous Aey17 mice but absent in all wild types, it was considered to be causative for the phenotype. 

The additional 49 bp led to a shift in the open reading frame and the protein to be truncated after 54 new amino acids (p.S110TfsX55; Fig. 3d ). It is predicted that the Ffg10 ^aey17 mutation destroys the Fgf10 protein after the first heparin-binding domain (Fig. 3e) . 

Effects of Lens Development in Homozygous Mutant Mouse Embryos on Fgf10

To understand in more detail the influence of Fgf10 on eye development in the mouse, we analyzed its expression by in situ hybridization (Fig. 4a) . At E12.5, it is present in the lens and in the retina. At E15.5, it is restricted in the lens to the epithelial cells and in the retina to the posterior parts of the inner and outer nuclear layers. At E17.5, it remains present in the lens epithelial cells, and in the retina it is restricted to the prospective photoreceptor cells. There is no major difference in the expression pattern and staining intensity between wild-type and homozygous Fgf10 ^aey17/aey17 mutants. 

Correspondingly, histologic analysis of the embryonic eyes of homozygous Fgf10 ^aey17/aey17 mutants showed no major effect on lens development at E12.5; the lens vesicle is formed properly, and the primary fiber cells are elongated (Fig. 5a) . Similarly, at E17.5, the eye is well organized, and the ocular tissues are properly differentiated. Only the lens nuclei do not show the characteristic distribution on the central lens area as is usually observed at this stage (Fig. 5b) . 

After birth, only heterozygous mutants survive. To identify any putative changes in the retina, Fgf10 ^+/aey17 mice (on the C3H-background, homozygous for the recessive Pde6b ^rd1 mutation) have been outcrossed to C57BL/6 mice (wild type for Pde6b). Histologic analysis of eyes heterozygous for Fgf10 ^+/aey17 demonstrated only minor changes. At P1, in the retinas of some mutant eyes (3 of 4), bulges were observed, but they could not be detected at P7. Moreover, the arrangement of the lens fiber cell nuclei is more diffuse in the mutants at P7 (data not shown). According to the histologic observations, no differences were detectable by funduscopy (data not shown). Moreover, functional analysis of the retina and the optic nerve by electroretinography and the virtual drum did not reveal any major difference between these and wild-type eyes (Table 2) . 

Discussion

The slit-eye phenotype as observed in the Aey17 mutant described here is at a first glance similar to the more common small-eye phenotype. However, detailed ophthalmic investigation demonstrated that the size of the mutant eye is not different from that of wild-type mice, and all other morphologic or functional parameters tested did not show any alteration. The only difference is the atrophy of the Harderian glands in the mutants. The Harderian glands synthesize lipids, pheromones, porphyrins, indoles, and saliva; it is discussed as involved in osmoregulation, immune response, photoprotection, and establishment of the retinal-pineal axis (for an overview, see Ref. ¹² and references therein). 

Recently, two mutants with a phenotype comparable to that of heterozygous Aey17 mice—PPP2CA transgenic mice ¹² and Muc1 null mice ¹³ —have been described. The PPP2CA transgenic mice express a human mutation (L309A) of the PPP2CA gene (coding for the protein phosphatase 2A catalytic subunit) in the Harderian gland, leading to its delayed postnatal development and hypoplasia. The authors showed that the morphology and function of the eye were not affected, but they reported a slit-eye phenotype (enophthalmos). The authors further indicated that this phenotype was a consequence of pronounced changes in proteins regulating cell adhesion. 

The Muc1 null mice ¹³ do not produce Mucin-1 in the Harderian gland; the resultant mutants have a slit-eye phenotype. In the Muc1-knockout mice, additional blepharitis and conjunctivitis are observed, most likely because of the missing protective effect of this protein on ocular epithelial surfaces. Both examples demonstrate that the loss of the Harderian gland (or at least essential proteins synthesized in this tissue) is able to cause a slit-eye phenotype. 

The mutant Aey17 was identified in an ENU screening program together with other small eye mutants, such as Aey11, Aey12, and Aey18. Among them, only Aey12 produced homozygous viable offspring and was characterized recently by a mutation in a novel gene, Gjf1, coding a connexinlike protein. ¹¹ As in Aey17, the Aey11 and Aey18 lines did not produce homozygous offspring, and Aey11 and Aey18 have been identified as mutations in the Pax6 gene. ⁹ However, linkage analysis mapped the novel Aey17 mutation to the distal end of mouse chromosome 13, and the candidate gene approach identified a mutation in the first intron of the Fgf10 gene (g.IVS1–2A>G). This mutation leads to the inclusion of 49 bp from intron 1 into the open reading frame of the Fgf10 transcript. Molecular analysis predicted the presence of 54 new amino acids together with 109 N-terminal amino acids of the wild-type sequence (p.S110TfsX55). The Fgf10 ^aey17 mutant is the first point mutation in the Fgf10 gene in the mouse, and the N-terminal part of the protein remains intact. 

Three other transgenic or knockout alleles of Ffg10 in the mouse have been reported (mgi database: http://www.informatics.jax.org; April 2009). Fgf10-knockout mice exhibit similar severe phenotypic characteristics, as do the Fgf10 ^aey17 mutants. The phenotype reported in the knockout mutants shows severe problems during embryonic development of limb and lung, ²⁵ loss of the lacrimal gland in the eye, ²⁶ formation of the inner ear, ²⁷ and cecal development. ²⁸ However, the slit-eye phenotype of the Fgf10 ^aey17 mutant is in contrast to the open eye lids observed in the Fgf10 knockout mutants at birth. ²⁹  

The Fgf10 ^aey17 slit-eye phenotype is similar to that of human disorders caused by mutations in the FGF10 gene (OMIM 602115). They have been associated with the dominant LADD syndrome (lacrimoauriculodentodigital syndrome; OMIM 149730) and dominant ALSG (OMIM 180920). The ALSG phenotype in humans has been characterized by irritable eyes and dryness of the mouth (xerostomia). In summary, ALSG was found in several point mutations of FGF10, including Arg80Ser, ³⁰ Gly138Glu, ³⁰ Lys137X, ³¹ Arg193X, ³² and a 53-kb deletion including exons 2 and 3. ³² In contrast, LADD syndrome is caused by other point mutations, including Cys106Phe ³³ and Ile156Arg ³¹ ; these human mutations are schematically shown in Figure 3e . However, there are some variations in the expressivity of the mutations. Milunsky et al. ³¹ reported a family with a mother with ALSG and her daughter with LADD (Lys137X). From the molecular point of view, the mouse Fgf10 ^aey17 is similar to the human 53-kb deletion, which interrupts the FGF10 gene at the same position as in the Fgf10 ^aey17 mouse mutant, leading to an intact first exon and most likely to a hybrid protein if the corresponding protein is stable. 

The irritability of the eye reported for ALSG is most likely caused by lacrimal gland aplasia, leading to dry eye syndrome and subsequently to recurrent infections. In general, clinically diagnosed dry eye syndrome has a prevalence of 0.4% to 0.5% and can have a substantial burden to the patients and a considerable economic impact. ³⁴ Given that the Fgf10 ^aey17 mutation is similar to the 53-kb deletion in human ALSG, it might be used as a new mouse model for the dry eye syndrome and probably allows a specific treatment. 

In conclusion, we report here the first point mutation in the mouse Fgf10 gene. Some effects of the mutation (major defects in limb development and perinatal lethality) are similar to those of the knockout mutation. However, the new slit-eye phenotype changes our knowledge of this gene, Fgf10, and opens the possibility of using the new mutant as a model for the dry eye syndrome, an important disease in the clinic. 

 Supported by Grants BMBF 01KW9923, BMBF 01GS0850, and LSHG-2006–037188, and in part by grants from the European Community to the German Mouse Clinic; the German Human Genome Project (EW, MHdA); and the National Genome Network (EW, IE, MHdA).

 Submitted for publication January 23, 2009; revised April 8, 2009; accepted June 24, 2009.

 Disclosure: O. Puk, None; I. Esposito, None; T. Söker, None; J. Löster, None; B. Budde, None; P. Nürnberg, None; D. Michel-Soewarto, None; H. Fuchs, None; E. Wolf, None; M. Hrabé de Angelis, None; J. Graw, 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.

 

Table 1.

 

Primers for PCR Reaction

Gene	Lab No.	Sequence (5′->3′)	Annealing Temperature (°C)	Length of PCR Product
Fgf10-L1	49175	CTTCTGATGAGACAATTTCCAGTGCC	57–62	711 bp
Fgf10-R1	49176	GAGTCATTGGTTGTACTGCATCCACC
Fgf10-Ex1-L1	50276	ACTTTCTCACGATTGAGAAGAACGGC	52–62	139/188 bp
Fgf10-Ex2-R1	50275	ATAGAGTTTCCCCTTCTTGTTCATGGC
Fgf10-Int-L1	50274	AACAGTTTTGGCTCAGTTTGGCTATCC	51–62	238 bp
Fgf10-Ex2-R1	50275	ATAGAGTTTCCCCTTCTTGTTCATGGC

Figure 1.

 

Ocular phenotype of the Aey17 mutants. (a) The slit-eye phenotype of an adult heterozygous Aey17 mouse (left) compared with a wild-type control (C3H/HeB; right). (b) Reduced size of a Harderian gland in a 4-month-old heterozygous Aey17 mutant (right) compared with a wild-type control (C3H/HeB; left). (c) Homozygous Aey17 mutants are not viable; arms and legs fail to develop.

Figure 2.

 

Pathology of the Harderian gland. (a) Overview of the Harderian gland in an adult wild-type mouse. Dark pigment concretions are seen in the alveolar lumens (H&E, 2.5×). (b–d) Harderian gland of an adult heterozygous mouse, represented by a fibrotic mass with abundant pigment deposits (b, H&E, 10×; c, Masson trichrome, 20×). Only a few epithelial cells with foamy, vacuolated cytoplasma are seen (d, H&E, 20×). These cells retain a positive reaction to cytokeratin AE1/AE3 (e, 40×). (f) Absence of immunoreactivity for the pan-epithelial marker cytokeratin AE1/AE3 in the Harderian gland of a heterozygous mouse (20×). In the wild-type animal, a diffuse positive reaction is seen (inset, 40×).

Figure 3.

 

Characterization of the causative mutation in the Aey17 mutant mice. (a) Haplotype analysis maps the Aey17 mutation between the markers D13Mit77 and D13Mit35, at the distal end of chromosome 13. (b) The candidate region between the two flanking markers D13Mit77 and D13Mit35 covers 2.4 MB with only a few genes; among them, Fgf10 is a strong candidate (ENSEMBL). (c) Sequence analysis of genomic DNA indicated an A/G heterozygous site in the mutant mice; the mutation does not affect a restriction site. (d) The A→G exchange at the end of intron 2 affects the highly conserved splice site; it changes the AG splice signal (green in the C3H sequence) to a GG splice signal (red in Fgf10^aey17/aey17). Sequence analysis of Fgf10 cDNA revealed the inclusion of 49 bp from the 3′ end of intron 1. Sequence comparison with the genomic DNA indicates that an upstream AG (green in the genomic Fgf10^aey17/aey17) is used as splicing signal. The additional 49 amino acids change the open reading frame, as indicated by the amino acid sequence above the cDNA sequence. Red/yellow: the altered amino acid sequence is highlighted. (e) Predicted domains of Fgf10 are given (according to ENSEMBL, modified). Red: transmembrane domain (TM). Green: heparin-binding domains (H). Numbers indicate the positions in the wild-type translation product. Arrow: position of the new mouse mutation. The C-terminal part is most likely replaced by a shorter but new amino acid sequence. Human mutations in the FGF10 gene are indicated below the scheme; the human protein is just one amino acid shorter than the mouse protein (208 aa vs. 209 aa).

Figure 4.

 

Expression analysis of Fgf10 in embryonic eyes. (a) Fgf10 expression in the eye is analyzed by in situ hybridization at E12.5, E15.5, and E17.5. At E12.5, it is present in the primary lens fibers, in the anterior lens epithelium, and in the retina. At E15.5, it is restricted to the epithelial cells, and in the retina it is restricted to two different layers (nerve fiber layer and outer nuclear layer). At E17.5, it remains present in the lens epithelial cells, and in the retina it is markedly expressed in the presumptive photoreceptor cells. (b) Expression pattern and staining intensity in homozygous mutant embryos are similar to those in wild type. (c) Sense probe is demonstrated for control (E15.5).

Figure 5.

 

Histology of embryonic eyes. (a) Histologic analysis of the embryonic eyes at E12.5. Wild-type eyes (left, from C3H mice) and eyes from Fgf10 ^aey17/aey17 mutants do not show major differences. (b) Histologic analysis of the embryonic eyes at E17.5. Wild-type eyes (left, from C3H mice) show the typical features of the cornea, the presence of lens fiber cell nuclei, and a two-layered retina. In contrast, the eye of the Fgf10 ^aey17/aey17 mutant is smaller at this stage of development, the lens epithelium of the mutants is thickened, and the lens fiber cells are swollen at the anterior part. C, cornea; L, lens; R retina; Scale bars indicate the original size.

Table 2.

 

Data of Electroretinography and Virtual Drum

	Wild-type	Heterozygous Aey17
ERG (500 cd/m2)
Mean a-wave amplitude (μV)	−7.3 ± 1.0	−7.9 ± 1.5
Mean implicit time (ms)	24.0 ± 3.5	26.5 ± 3.1
ERG (12500 cd/m2)
Mean a-wave amplitude (μV)	−27.5 ± 11.9	−36.5 ± 14.9
Mean implicit time (ms)	23.5 ± 2.3	23.4 ± 1.3
ERG (500 cd/m2)
Mean b-wave amplitude (μV)	157.8 ± 27.3	143.6 ± 40.0
Mean implicit time (ms)	100.5 ± 10.2	93.6 ± 19.2
ERG (12500 cd/m2)
Mean b-wave amplitude (μV)	178.2 ± 52.4	192.7 ± 32.8
Mean implicit time (ms)	70.2 ± 8.5	64.5 ± 10.7
Virtual optokinetic drum
Median threshold (cyc/deg)	0.414	0.427

 ERG, n = 4; virtual optokinetic drum, n = 8.

The authors thank Erika Bürkle, Jan Enke, Maria Kugler, Elenore Samson, and Monika Stadler for expert technical assistance; Utz Linzner (Helmholtz Center Munich, Institute for Experimental Genetics) for providing oligonucleotides; and Jack Favor (Helmholtz Center Munich) for critical reading of the manuscript.