Overexpression and Properties of Wild-Type and Tyr437His Mutated Myocilin in the Eyes of Transgenic Mice

Abstract

purpose. To obtain experimental in vivo information on the functional properties of myocilin for aqueous humor (AH) outflow and to study in vivo the processing of mutated Tyr437His myocilin. Myocilin is a secreted glycoprotein that is mutated in some forms of primary open-angle glaucoma (POAG), and patients with the Tyr437His mutation have severe phenotypes.

methods. The chicken βB1-crystallin promoter was used to overexpress wild-type human myocilin and mutated Tyr437His myocilin in the lenses of transgenic mice. Expression of transgenic mRNA was monitored by Northern blot analysis and in situ hybridization. The localization and secretion of transgenic myocilin was investigated by Western blot analysis and light and electron microscopy. Intraocular pressure (IOP) was measured by anterior chamber cannulation.

results. Two independent lines were established with each of the constructs that showed a strong expression of transgenic mRNA in their lenses. Transgenic expression resulted in a 4.7 ± 1.8-fold increase of secreted normal myocilin in mouse AH, compared with its concentration in human AH. Immunoreactivity for transgenic myocilin was observed along the surfaces of lens and corneal endothelium, and in the chamber angle. At 12 weeks of age, the ultrastructure of the trabecular meshwork in mice expressing normal myocilin was not different from that of control eyes, and IOP of transgenic animals did not significantly differ from that of control littermates. In contrast, mutated Tyr437His myocilin was not secreted from lens fibers, but accumulated in dilated cisterns of rough endoplasmic reticulum. Although no structural changes were observed in lenses of animals expressing normal myocilin, lenses with Tyr437His expression developed nuclear cataracts, completely lost transparency, and eventually ruptured. Structural changes in lenses of Tyr437His expressing mice were correlated with a significant increase in IOP.

conclusions. The results do not support the concept that increasing amounts of myocilin in the outflow tissues obstruct the system and directly cause an increase in outflow resistance. Mutated Tyr437His myocilin is not secreted in vivo and causes severe alterations of cellular structure and function. A comparable mechanism may cause POAG in patients with myocilin mutations.

Glaucoma, a major cause of blindness worldwide, ¹ is characterized by a gradual loss of axons in the optic nerve head. The major risk factor for glaucoma is an intraocular pressure (IOP) that is too high for the health of the optic nerve head. ^{2 3 4} IOP critically depends on the resistance to aqueous humor (AH) outflow in the trabecular meshwork (TM). In primary open-angle glaucoma (POAG), IOP is elevated because of an abnormally high outflow resistance in the TM. ⁵ Mutations in the myocilin gene have been found in families with autosomal-dominant juvenile POAG and some forms of adult-onset POAG. ^{6 7} Myocilin is a 55- to 57-kDa glycoprotein characterized by a coiled coil domain near the amino terminus and an olfactomedin domain near the carboxyl terminus. ⁸ Patients with myocilin mutations may have higher IOP and a more severe course of disease than those with other forms of POAG, suggesting that the major site of action of mutated myocilin is the TM. ^{8 9} There is evidence from a considerable number of in vitro studies that myocilin with glaucoma-causing mutations is not secreted from TM cells, but accumulates in the rough endoplasmic reticulum (rER). ^{10 11 12} Retention of mutated myocilin in cultured TM cells leads to abnormal cell morphology and cell death. ^{11 12} At present, it is unclear whether such a scenario also occurs in the TM of patients with POAG with myocilin mutations, as histopathological in situ data are lacking. 

Despite considerable research efforts, there is also uncertainty as to the function of normal myocilin. Significant expression of myocilin is found in the tissues of the anterior eye ^{13 14} and in Schwann cells of peripheral nerves. ¹⁵ In the anterior eye, the expression of myocilin is extremely high in the TM ^{13 16 17} suggesting a role for the modulation of AH outflow, the major function of the TM. Secreted myocilin is present in the AH, suggesting that it passes through the outflow pathways. ^{10 18 19 20} It has been hypothesized that secreted myocilin plays a direct role in modulating the hydrodynamic resistance in the TM and that elevated amounts of myocilin lead to obstruction of the outflow system. ²¹ This hypothesis is supported by several observations and experiments: (1) Dexamethasone induces expression of myocilin in cultured TM cells and perfusion organ cultures in a time-dependent manner, similar to the time course for the development of steroid-induced ocular hypertension and glaucoma. ²² (2) An increase in myocilin immunoreactivity has been observed in the TM of patients with POAG. ²³ (3) Recombinant myocilin is very effective at blocking polycarbonate filters with a pore size similar to that of the TM. ²⁴ In addition, myocilin in the AH is tightly bound to polycarbonate filters that become obstructed after perfusion with AH. ¹⁹ (4) Perfusion of human anterior segments in organ culture with partially purified recombinant myocilin from a bacterial expression system increases outflow resistance by 94%. ²⁵ Still, comparable perfusion experiments with a myocilin fragment purified from a eukaryotic expression system did not show significant effects on outflow facility. ²⁴  

Our goal in this study was to determine the in vivo effects of mutated myocilin and elevated amounts of normal myocilin in the eye. To this end, we have generated transgenic mice that overexpress mutated and wild-type myocilin under control of a lens-specific promoter. Increasing amounts of wild-type myocilin in the AH of transgenic mice did not obstruct the outflow system and did not cause an increase in IOP. Mutated Tyr437His myocilin, which is associated with a severe POAG phenotype in humans, ²⁶ is not secreted in vivo, but is retained in the rER, causing severe alterations of cellular structure and function. 

Materials and Methods

All procedures in this study conformed to the tenets of the Declaration of Helsinki, the National Institutes of Health Guidelines on the Care and Use of Animals in Research, and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Generation of Transgenic Mice

The full-length, 1811-bp MYOC cDNA was obtained and cloned into plasmid pERTI47-2, as described previously. ¹⁶ The cDNA was released by using the flanking BamHI sites and cloned into the compatible BglII sites of plasmid pER17-5. ²⁷ This resulted in plasmid pβB1myoc, which contained the chicken βB1-crystallin promoter, followed by a 479-bp intron from plasmid pOPI3CAT and the myocilin cDNA. Thymidine kinase (TK) polyA sequences provided an additional polyadenylation signal (Fig. 1A) . The orientation and nature of all fragments in plasmid pβB1myoc were confirmed by automated sequencing with fluorescent dideoxynucleotides (model 310; Applied Biosystems, Foster City, CA). 

For cloning of βB1-crystallin-Tyr437HisMYOC mice, a myocilin cDNA fragment was amplified by PCR from plasmid pERTI47-2 using the primer pair 5-CTTAGACCTGGAGGCCACC-3 and 5-GCCTGCTCCCCCCAGGAGCC-3 and cloned into the SrfI site of plasmid pPCR-Script (Stratagene, La Jolla, CA) to obtain pERTI36-3. Site-directed mutagenesis to insert the Tyr437His amino acid substitution was achieved by PCR, using the primer pair 5-AAGCAGGCCTCTGGGTCATTTACAGCACCGATGAGGCCAAAGGT-3 (StuI) and 5-TCTGCTGAGGTGTAGCTGCTGACGGTGTGCAAGGTGCCACAGAT-3 (Bpu10I) with StuI and Bpu10l restriction sites (underlined) at the ends to facilitate cloning. The resultant PCR product was cloned between the StuI and Bpu10I sites of plasmid pERTI36-3 to obtain plasmid pERTI49-3. Site directed mutagenesis was confirmed by automated sequencing. pERTI49-3 was digested with StuI and SstI, and the resultant fragment was cloned between StuI/SstI sites of pERTI47-2. The resultant plasmid pERTI50-5 contained the complete open reading frame of myocilin with a T→C point mutation resulting in the Tyr437His amino acid exchange, but lacked 3′-untranslated region (UTR) sequences and the polyA termination sites of MYOC. The Tyr437His mutated MYOC cDNA was released by using the flanking BamHI sites and cloned into the compatible BglII sites of plasmid pER17-5 as described earlier. The resultant plasmid pβB1mutmyoc was again analyzed by automated sequencing. Constructs for microinjection were released from plasmids pβB1myoc and pβB1mutmyoc by digestion with XhoI and XbaI followed by gel electrophoresis. FVB/N transgenic mice were generated at the National Eye Institute Centralized Transgenic Facility as described by Wawrousek et al. ²⁸  

Transgenic Mice

Potential βB1-crystallin-MYOC or βB1-crystallin-Tyr437HisMYOC transgenic mice were screened by isolating genomic DNA from tail biopsies and testing for transgenic sequences by Southern blot hybridization and/or by the use of PCR. Southern hybridization was performed on each transgenic family by using 10 μg of EcoRI-digested genomic tail DNA to determine the number of integration sites and the integrity of the transgene. For PCR analysis, primers were used that span from promoter sequences to the intron of the transgene. The sequences of the primers were 5-ACACTGATGAGCTGGCACTTCCATT-3 and 5-TGTTGGCTACTTGTCTCACCATTGTA-3. A 506-bp DNA fragment was amplified and visualized by agarose gel electrophoresis and ethidium bromide staining. The thermal cycle profile was denaturation at 94°C for 50 seconds, annealing at 50°C for 50 seconds, and extension at 72°C for 1 minute for 30 cycles. 

Transgenic mice were housed under standardized conditions of 62% air humidity and 21°C room temperature. Feeding was ad libitum. Animals were kept at a 12-hour light–dark cycle (6 AM to 6 PM). Before enucleation of the eyes, mice were anesthetized with CO₂ and killed by atlanto-occipital dislocation. 

RNA Analysis

To detect transgenic mRNA by digoxigenin (DIG)-labeled riboprobes, PCR products were amplified by using plasmid pβB1myoc as the template. To generate an antisense probe for transgenic mRNA, sequences of the T7-promotor were added to the downstream primer. The sequence of the primer pairs were 5-CCTCCTCCGAGACAAGTCAG-3 and 5-ATCGATAATACGACTCACTATAGGGCCCTGCATAAACTGGCTGAT-3. The PCR-product contained 502 bp of human MYOC sequence and was gel-purified by using a PCR kit (GFX; Amersham, Amersham, UK), according to the manufacturer’s instructions. The antisense RNA probe was labeled with DIG-11-UTP by using T7-polymerase according to the manufacturer’s instructions (Roche, Mannheim, Germany). 

For Northern blot analysis, total RNA was isolated from ocular tissues (TRIzol; Invitrogen, Karlsruhe, Germany) according to the manufacturer’s instructions. RNA from transfected 293 EBNA MYOC cells ²⁴ was used as a positive control. The molecular weight of MYOC mRNA from 293 EBNA MYOC cells slightly differed from that of transgenic animals, as a C-terminal sequence of six histidine residues (HisTag) had been added, and the MYOC signaling peptide had be exchanged with that of BM40 (osteonectin) to facilitate secretion of recombinant protein. ²⁴ RNA was separated on a 1% agarose gel containing 6% formaldehyde and blotted on a positively charged nylon membrane (Roche). After transfer, the blot was cross-linked (UV Stratalinker 1800; Stratagene). Prehybridization was performed for 1 hour at 68°C and hybridization overnight at 68°C (DIG EasyHyb-buffer; Roche). Membranes were washed 5 minutes with 2× SSC and 0.1% SDS at room temperature and 15 minutes with 0.2× SSC and 0.1% SDS at 70°C. For detection of the hybridization signals, membranes were blocked 30 minutes at room temperature in blocking solution (1% blocking reagent, 0.1 M maleic acid, and 0.15 NaCl; pH 7.5) and incubated in anti-DIG-alkaline phosphatase Fab-fragments diluted 1:10,000 (all Roche). After washing the membranes two times for 15 minutes in 0.1 M maleic acid, 0.15 M NaCl (pH 7.5), and 0.3% Tween 20, chemiluminescence detection was performed (CDP-Star, 1:100; Roche). Membranes were then exposed (Lumi-Imager workstation; Roche) and intensities of hybridization signals were determined by using the accompanying software (Lumi-Analyst; Roche). To monitor the integrity of RNA, the relative amounts of RNA loaded on the gel, and the efficiency of transfer, membranes were stained with methylene blue. 

For in situ hybridization, heads of embryonic day (E)13.5 mice were incubated for 4 hours in phosphate-buffered saline (PBS) containing 4% paraformaldehyde (pH 7.5) at 4°C, washed overnight in PBS containing 30% sucrose at 4°C, frozen, and stored in liquid nitrogen. Cryosections were cut on a cryotome and dried at 40°C overnight. Sections were treated with proteinase K (1 μg/mL), postfixed in PBS containing 4% paraformaldehyde, and acetylated with Tris acetate EDTA (TEA) buffer (pH 8.0) containing 0.25% acetic anhydride. Prehybridization was performed in 4× SSC containing 50% deionized formamide at 37°C. For hybridization, sections were incubated in 40% deionized formamide, 10% dextran sulfate, 1× Denhardt’s solution, 4× SSC, 10 mM dithiothreitol (DTT), 1 mg/mL yeast tRNA, and 1 mg/mL salmon sperm DNA at 55°C overnight with antisense RNA probes added to a concentration of 0.1 μg/mL. The slides were rinsed twice for 15 minutes in 2× SSC at 50°C followed by two 15-minute washes in 1× SSC at 50°C. Nonhybridized, single-stranded RNA probe was removed by RNase A digestion (20 μg/mL) in 500 mM NaCl, 10 mM Tris, and 1 mM EDTA (pH 8.0) for 30 minutes at 37°C, followed by two 30-minute washes in 0.1× SSC containing 50% formamide at 50°C. Slides were blocked for 30 minutes in 100 mM Tris-HCl (pH 7.5), 150 mM NaCl, 0.1% Triton X-100, and 1% normal sheep serum. Sheep anti-DIG-alkaline phosphatase antibodies (1:500; Roche) were added, and slides were incubated for 2 hours in a humid chamber. After two 10-minute washes with 100 mM Tris-HCl (pH 7.5) and 150 mM NaCl, slides were incubated for 2 hours in nitro blue tetrazolium/5-bromo-4-chloro-3-indoxyl phosphate/ (NBT/BCIP) color solution (Roche). After adding stop solution (Roche), slides were rinsed in sterile water and mounted. 

Western Blot Analysis

Ocular tissues were lysed in SDS sample buffer. Proteins (5 μg/lane) were separated by polyacrylamide gel electrophoresis using a 5% SDS-polyacrylamide stacking gel, and 8% to 15% or 12% SDS-polyacrylamide separating gels. Human recombinant myocilin obtained from transfected 293 EBNA MYOC cells, as described previously, ²⁴ was used as a positive control. Most gels were run under reducing conditions, although selected gels were run under nonreducing conditions. After electrophoresis, the proteins were transferred with semidry blotting (Bio-Rad, Hercules, CA) on a polyvinylidene difluoride membrane (Roche). The membrane was incubated with PBS containing 0.1% Tween 20 (PBST, pH 7.2) and 5% bovine serum albumin (BSA) for 1 hour. An anti-human myocilin antibody (diluted 1:1000, Santa Cruz Biotechnology, Inc., Santa Cruz) was added overnight at 4°C. After washing with PBST, alkaline phosphatase-conjugated donkey anti-goat IgG (diluted 1:10,000, Santa Cruz Biotechnology, Inc.) was added for 30 minutes. Alkaline phosphatase was visualized by using chemiluminescence (CDP-Star; Roche) and the workstation (Lumi-Imager workstation; Roche). Exposure times ranged between 1 and 5 minutes. 

Dot Blot Analysis

One-microliter samples of AH from postnatal day (P)21 βB1-crystallin-MYOC mice and wild-type littermates and six different human AH samples were dotted onto a nitrocellulose membrane and allowed to air dry. The nitrocellulose membrane was stained with a fluorescent protein gel stain (SYPRO Ruby; Molecular Probes Inc., Eugene, OR) according to the manufacturer’s instructions. Briefly, the nitrocellulose was incubated in 7% acetic acid and 10% methanol for 15 minutes and then washed three times with deionized H₂O. The nitrocellulose was stained with the fluorescent stain for 15 minutes, washed quickly with deionized H₂O, and then the protein spots on the paper were visualized with an imaging station (Image Station 440; Eastman Kodak, Rochester, NY) using UV illumination. The paper was then blocked (Superblock; Pierce Biotechnology Inc., Rockford, IL) and processed like a Western blot. The primary myocilin antibody raised in chicken (N-[KSELTEVPASRILKES]-C) was diluted 1:10,000, and the membrane was incubated overnight. The secondary antibody was diluted 1:50,000. Development was performed with a chemiluminescence kit (Western Lightning; Perkin-Elmer Life Sciences, Wellesley, MA). The volumes of both the SYPRO Ruby spots and the antibody spots were quantitated on computer (Image Quant software; Molecular Dynamics, Sunnyvale, CA). A standard curve was obtained from the BSA-stained spots. The volume of the myocilin stain was divided by the protein present in that spot. A mean for this value from the human AH samples was set at 1.0 and the other spots were normalized to this value. Duplicate blots were run using the chicken anti-myocilin antibody as well as an additional blot that used the rabbit anti-myocilin antibody previously described. ²⁹ Similar results were obtained with all three blots. 

Light and Electron Microscopy

Embryos were obtained from timed matings, with noon of the day the vaginal plug was discovered designated as embryonic day (E)0.5 of gestation. For light microscopy and/or immunohistochemistry heads (E13.5–E14.5) or eyes (E15.5–P21, and 12 weeks of age) were fixed for 24 hours in 4% paraformaldehyde and embedded in paraffin or in methacrylate (Technovit 7100; Hereaus Kulzer GmbH, Wehrheim, Germany), according to standard protocols. For electron microscopy, eyes were placed in Ito’s fixative ³⁰ for 24 hours. In older embryos, the posterior half of the eye was pierced with a fine needle. Samples were washed overnight in cacodylate buffer, postfixed with OsO₄, dehydrated, and embedded in Epon (Roth, Karlsruhe, Germany). One-micrometer, semithin sections were stained with toluidine blue. Ultrathin sections were stained with uranyl acetate and lead citrate and viewed by electron microscope (model EM 902; Carl Zeiss Meditec, Oberkochen, Germany). 

Immunohistochemistry

Staining for myocilin was performed by using rabbit ^{29 31} or chicken ³² anti-myocilin antibodies, as described previously. Before overnight incubation at 4°C with the primary antibody, paraffin sections were incubated in nonfat dry milk solution ³³ for 30 minutes. The primary antibody was removed from the sections by three washes (10 minutes each) in PBS, and the sections were treated for 1 hour with biotinylated secondary antibodies specific against the respective primary antibody (Vector Laboratories, Burlingame, CA). After three washes in PBS, the sections were covered with fluorescein isothiocyanate (FITC) streptavidin (Vector Laboratories) for 1 hour, washed again, mounted in fluorescent mounting medium, and viewed with a fluorescence microscope (Orthoplan; Leitz, Wetzlar, Germany). 

TUNEL Labeling

Apoptosis was detected in transgenic mice and wild-type littermates at 12 weeks of age by the TUNEL assay (Apoptosis Detection System with fluorescein; Promega Corp., Madison, WI). Paraffin-embedded sections were treated with proteinase K (20 mg/mL) for 10 minutes at room temperature, washed, fixed in 4% methanol-free formaldehyde solution for 5 minutes, and covered with equilibration buffer. TdT (terminal deoxynucleotidyl transferase)-buffer containing nucleotide mix was applied to the sections for 60 minutes at 37°C in a moist chamber. The sections were washed and counterstained with propidium iodide. 

IOP Measurements and Aqueous Humor Collection

IOP was measured invasively in the eyes of 12-week-old transgenic mice (body weight, 27–30 g) and wild-type littermates according to the method described by John et al. ³⁴ Measurements were performed at the same time of day. The animals were deeply anesthetized by intramuscular injection of a mixture of ketamine as hydrochloride (100 mg/kg; Parke-Davis, Freiburg, Germany) and xylazine (5 mg/kg; Bayer Leverkusen, Leverkusen, Germany) and placed on a surgical platform. The eye was viewed under a dissecting microscope, and the 30-gauge microneedle tip was placed inside a drop of PBS on top of the eye. At this point, the pressure reading was zeroed. The tip of the microneedle was manually inserted into the anterior chamber with the IOP recorded continuously. The IOP values usually reached a plateau after 1 second of higher IOP due to the needle insertion. If the plateau was constant for 1 minute, the eye was given some gentle extra pressure from outside to confirm microneedle patency. This manipulation resulted in some IOP peaks that returned to the prior plateau level. The microneedle was then withdrawn from the anterior chamber; a rapid return of the pressure to zero was necessary for the inclusion of data. The original measurements were calibrated in cmH₂O and converted to millimeters of mercury (mm Hg). For statistical analysis, a Student’s t-test was performed. 

For the collection of mouse AH, 3-week-old transgenic mice and wild-type littermates were deeply anesthetized and placed on a surgical platform. The eye was observed under a dissecting microscope while the AH was collected by inserting a fine glass microneedle into the anterior chamber of the eye. The AH samples were frozen immediately at −20°C. 

Human aqueous humor was obtained at the time of cataract removal. An institutionally approved human protocol governed the acquisition of the human samples, and informed consent was obtained from all patients. 

Results

βB1-Crystallin-MYOC Transgenic Mice

Transgenic Overexpression of Normal Myocilin.

Transgenic mice that expressed a 1.8-kb human MYOC cDNA driven in the lens by the chicken βB1-crystallin promoter were generated. The chicken βB1-crystallin promoter (−434/+30) has been shown to drive lens-specific expression in transgenic mice and to be active in both primary and secondary lens fibers from E12.5 until adulthood in a copy number and position-independent manner. ^{27 35 36} A 479-bp intron from plasmid pOPI3CAT was placed between the chicken βB1-crystallin promoter and the MYOC cDNA, which contained the MYOC polyA termination sites. Thymidine kinase (TK) polyA sequences provided an additional polyadenylation signal (Fig. 1A) . The βB1-crystallin-MYOC fragment was injected into pronuclear stage FVB/N embryos to obtain transgenic mice. Three initial founder animals were identified by PCR and designated 9-3, 9-8, and 9-16, respectively. Three independent transgenic families were produced with each founder animal, and Southern blot hybridization was performed to confirm that transgenic animals from each family contained a single transgene insertion site. In each of the lines, 8 to 10 copies of the transgene were present. The size of transgenic eyes did not obviously differ from that of wild-type littermates. The cornea and lens of all transgenic mice appeared to be normal and translucent on eye opening. 

Transgenic mRNA expression in each of the three lines was analyzed by Northern blot hybridization. At P1, total RNA was collected from lenses of all three transgenic lines and wild-type littermates. By using a probe specific for transgenic mRNA, we detected positive hybridization signals in RNA from lenses of the transgenic lines 9-8 and 9-16, whereas no signal was observed in RNA from line 9-3 or wild-type littermates (Fig. 1B) . In high-resolution gels, four bands were observed in lines 9-8 and 9-16 (Fig. 1B) . The sizes of the two faster migrating bands corresponded to that of human TM MYOC mRNA, which has been shown to migrate between 1.8 and 2.4 kb. ^{16 37} It was concluded that the two faster migrating bands in transgenic mRNA resulted from the use of both MYOC and TK polyA sequences. In addition, two slower migrating bands were observed, which probably corresponded to preprocessed mRNA. To determine whether the chicken βB1-crystallin promoter drives the expression of MYOC lens-specific mRNA, similar to other transgenes that are expressed under control of this promoter, ^{27 35} the expression of MYOC mRNA in lenses of transgenic animals was compared with the expression in the rest of the eye. Northern blot experiments in transgenic lines 9-8 (Fig. 1C)and 9-16 (not shown) gave similar results. In P1 animals, an intense expression of transgenic mRNA was observed in RNA from the lens, but not in RNA from the remainder of the eye (Fig. 1C) . In wild-type animals, no signal for MYOC was observed, neither in RNA from the lens nor in that from the remainder of the eye (Fig. 1C) . In contrast, a positive signal was observed in mRNA from 293 EBNA cells that had been transfected by a vector expressing the human MYOC cDNA (EBNA MYOC), ²⁴ but not in control EBNA cells (Fig. 1C) . Only one band was observed in RNA from EBNA MYOC cells, consistent with the presence of a single polyA termination site in the expression vector that had been used for transfection of the cells. ²⁴ The band in EBNA MYOC cells comigrated with the fastest migrating band in RNA from transgenic animals. To confirm lens-specific expression in mice from line 9-8, in situ hybridization was performed at E13.5 in transgenic and control animals by using antisense probes specific for transgenic mRNA. Transgenic mRNA expression was specifically localized to elongating lens fiber cells of transgenic animals and was absent in lens epithelium (Fig. 1D) . 

To study whether transgenic mRNA was effectively translated, we performed Western blot analysis on lens proteins from P1 animals by using antibodies specific against myocilin. In lenses from animals of lines 9-8 and 9-16, a weak band was observed that comigrated with recombinant human myocilin (Fig. 2A) . In contrast, no immunoreactivity was observed in lens proteins from line 9-3 or wild-type littermates. By P21, no bands were observed by Western blot analysis of lens proteins from line 9-16. There was, however, a distinct positive band in the AH of the same animals, but not in lenses or AH of wild-type littermates (Fig. 2B) . Essentially similar results were obtained for line 9-8 (not shown). 

To obtain a measure for the relative amounts of transgenic myocilin, AH of four 9-16 animals was compared with equal amounts of AH from six human donors by Western dot blot analysis. AH from human donors was chosen, as transgenic animals have been genetically modified to overexpress human myocilin. Quantitative analysis of antibody spots showed 4.7 ± 1.8 more human myocilin in the AH of transgenic mice than in that of human donors (Fig. 2C) . Comparable results were obtained by comparing human and pooled 9-16 AH by one-dimensional Western blot analysis (Fig. 2D) . In addition, comparison with different amounts of recombinant human myocilin indicated that transgenic myocilin was present in the AH of transgenic 9-16 animals at a concentration of approximately 0.2 ng/μL (Fig. 2D) . Both recombinant myocilin and myocilin in the AH of human donors and transgenic mice resolved as a doublet on high-resolution, reducing gels (Fig. 2D) . Under nonreducing conditions, bands of higher molecular weight were detected consistent with the formation of myocilin multimers (not shown). 

Structural and Functional Analysis of βB1-Crystallin-MYOC Transgenic Mice.

The localization of transgenic myocilin was studied in anterior eyes of P14 transgenic 9-16 animals by immunohistochemistry. No positive signal was detected in lenses of transgenic animals and wild-type littermates (Figs. 3A 3B) . Lenses of transgenic animals showed a thin, immunoreactive layer on their surfaces, indicating that lenses were partially covered with transgenic myocilin (Fig. 3B) . A similar immunoreactive layer was observed covering parts of the inner surface of the corneal endothelium (Fig. 3D) . Corneal endothelial and stromal cells, cells of the ciliary body, and cells of the chamber angle were positive in both transgenic and wild-type animals corroborating previous findings (Fig. 3C 3D 3E 3F) . ³¹ In addition, in transgenic animals, a homogenous mass with strong immunoreactivity for myocilin was observed in the chamber angle covering the inner parts of the TM (Fig. 3F) . This material was not observed in control animals, indicating that transgenic myocilin accumulated in the chamber angle. Immunoreactivity for myocilin within the TM proper was not different between transgenic and wild-type animals (Figs. 3E 3F) . 

Conventional light microscopy of the anterior eye did not show obvious structural differences between transgenic mice and wild-type littermates at 12 weeks of age. In both groups of animals, the structure of ciliary body, iris, and TM was essentially normal (Figs. 4A 4B) . The same was true of the bow region of the lens in transgenic animals, where elongation of lens fibers and activation of the βB1-crystallin promoter starts (Fig. 4D) . At higher magnification, the extracellular spaces in the TM of transgenic animals appeared to be optically empty (Fig. 4C) . Electron microscopy of transgenic animals showed that the cells of the TM were structurally intact (Fig. 4E) . In both the inner corneoscleral and uveal parts of the TM and the outer juxtacanalicular part close to the endothelial lining of Schlemm’s canal, the extracellular spaces that serve as flow pathways for the AH, were not filled with unusual amounts of extracellular material, but were largely empty (Figs. 4E 4F) . In the juxtacanalicular part, some extracellular fine fibrillar material was observed, the amount of which did not differ from that in control animals (Fig. 4F) . 

At 12 weeks of age, IOP was 8.8 ± 3 mm Hg (mean ± SD, n = 18) in transgenic animals, and 9.9 ± 2.2 mm Hg (n = 16) in wild-type littermates (Fig. 5) , a statistically nonsignificant difference. 

βB1-Crystallin-Tyr437HisMYOC Transgenic Mice

For the generation of βB1-crystallin-Tyr437HisMYOC transgenic mice, a mutated 1.59-kb MYOC cDNA was used, in which the Tyr437HisMyoc mutation had been inserted by site-directed mutagenesis. The mutated cDNA lacked 3′-untranslated sequences and the MYOC polyA termination sites. Otherwise, the structure of the transgenic construct was comparable to that used for the generation of βB1-crystallin-MYOC transgenic mice (Fig. 6A) . 

After microinjection and embryo transfer, three initial founder animals were identified by PCR and designated 8-5, 8-6, and 8-13, respectively. Three independent transgenic families were produced with each founder animal and Southern blot hybridization was performed to confirm that transgenic animals from each family contained a single transgene insertion site. In lines 8-5 and 8-6, 20 copies of the transgene were present, whereas line 8-13 contained only 3 copies. The size of transgenic eyes did not obviously differ from that of wild-type littermates. 

Transgenic mRNA expression in each of the three lines was analyzed by Northern blot hybridization at P1. Positive hybridization signals were detected in RNA from lenses of transgenic lines 8-5 and 8-6, whereas no signal was observed in RNA from line 8-13 (Fig. 6B) . When the expression of mutated MYOC mRNA in lenses of transgenic animals was compared with the expression in the rest of the eye, a strong signal was observed in RNA from lenses of transgenic animals, but not in RNA from eye tissues without the lens or in RNA from wild-type littermates (Fig. 6C) . The band, which was observed in transgenic lens RNA comigrated with that observed in RNA from EBNA MYOC cells, which were used as positive control (Fig. 6C) . Similar results were obtained in transgenic lines 8-5 (Fig. 6C)and 8-6 (not shown). 

In Western blot analysis of lens proteins from P21 βB1-crystallin-Tyr437HisMYOC transgenic mice, a distinct positive band was observed in lens proteins, whereas only a weak band was detectable in the AH of the same animals (Fig. 6D) . This result differed markedly from that observed in βB1-crystallin-MYOC mice, which overexpressed normal myocilin (Fig. 2B) . 

Structural Analysis of βB1-Crystallin-Tyr437HisMYOC Transgenic Mice

Light microscopy of P21 βB1-crystallin-Tyr437HisMYOC transgenic mice showed multiple vesicles in the lens fibers at the bow region of transgenic lenses (Fig. 7B) . Such vesicles were not observed in cells of the lens epithelium of transgenic animals (Fig. 7B) , nor in lenses of wild-type littermates (Fig. 7A)or transgenic mice with overexpression of normal myocilin (Fig. 4D) . By electron microscopy, the vesicles were found in the perinuclear cytoplasm of lens fibers and were filled with electron-dense granular material (Fig. 7D) . In contrast, in wild-type littermates, perinuclear areas of lens fibers contained cisterns of rER that appeared to be of normal size (Fig. 7C) . Higher magnification of perinuclear vesicles in transgenic animals revealed that the vesicles were surrounded by a membrane that contained ribosomes and confirmed their origin from rER cisterns (Fig. 7F) . Membrane-free, cytoplasmic inclusion bodies or aggresomes were not observed. Immunocytochemistry with antibodies specific for myocilin showed strong positive immunoreactivity in lens fiber vesicles of transgenic animals, and confirmed that the vesicles were caused by an accumulation of Tyr437His mutated myocilin in the rER (Figs. 7G 7H) . 

The accumulation of mutated myocilin in lens fibers in the bow region did not cause other apparent structural changes in the fibers or apoptotic cells death, as TUNEL staining of the cells was essentially negative (not shown). In contrast, marked structural alterations were present in nuclear lens fibers, as βB1-crystallin-Tyr437HisMYOC transgenic mice developed a nuclear cataract that caused loss of lens transparency (Fig. 8A) , whereas lenses of wild-type littermates remained optically clear (Fig. 8B) . Nuclear cataract in βB1-crystallin-Tyr437HisMYOC transgenic mice eventually led to rupture of the lens at its posterior pole, which was observed in most of the animals at P21 (Fig. 8C) . 

At 12 weeks of age, IOP was 10.6 ± 2.9 mm Hg (mean ± SD, n = 14) in βB1-crystallin-Tyr437HisMYOC transgenic animals and 7.7 ± 1.6 mm Hg (n = 13) in wild-type littermates (Fig. 9) , a statistically significant difference (P < 0.004). 

Discussion

We conclude that increasing amounts of myocilin in the AH have little or no effect on outflow facility. This conclusion is based on the substantial increase of myocilin in the AH of βB1-crystallin-MYOC transgenic mice, the presence of normal IOP in βB1-crystallin-MYOC transgenic mice, and the lack of any detectable structural alterations in the AH outflow pathways of βB1-crystallin-MYOC mice. 

The results of this study were made possible by the fact that myocilin was produced in substantial amounts in transgenic lenses and secreted readily into the AH. The high and lens-specific expression of transgenic mRNA driven by the chicken βB1-crystallin promoter is consistent with other reports, in which the same promoter was used for the expression of transgenes in the mouse eye. ^{27 35} Transgenic mRNA was translated effectively, which resulted in an approximately fivefold increase of myocilin in the AH of transgenic mice, compared with the concentration of myocilin in human AH. The increase of myocilin in the AH of βB1-crystallin-MYOC mice would be even higher when compared with the amount of myocilin in the AH of wild-type mice, in which no myocilin was detected, although antibodies were used that have been shown to label mouse myocilin. ³¹ Still, βB1-crystallin-MYOC mice express human myocilin and we cannot exclude the possibility that the antibodies, which have been generated against human peptide sequences, may be more sensitive in detecting human myocilin than mouse myocilin. 

It is reasonable to assume that transgenic myocilin in the AH of βB1-crystallin-MYOC mice was drained into the chamber angle thereby increasing the local concentration of myocilin in the TM. Our data indicate that transgenic myocilin in the TM did not cause changes in TM structure or function that may have resulted in a decrease of TM outflow facility, leading to an increase in IOP. The hypothesis that an increase of myocilin in the TM may cause a decrease in TM outflow facility has been put forward based on the observation that the expression of myocilin in cultured TM cells is increased after treatment with dexamethasone. ^{21 22} This hypothesis is supported by the observation that the time course of the steroid-induced myocilin synthesis in TM cells in vitro parallels the increase in IOP in patients with ocular hypertension due to treatment with steroids. ^{38 39 40} Clearly, our data do not support such a hypothesis, but rather indicate that the amount of myocilin in the TM does not directly correlate with TM outflow resistance. We cannot exclude, though, that with steroid treatment, more myocilin is synthesized in the TM than the substantial amounts that were generated in transgenic βB1-crystallin-MYOC mice in our study, but we regard such a scenario as unlikely. Alternatively, the mouse TM may be less sensitive to the presence of higher amounts of myocilin than the human TM. Still, the available data on the structural similarities between mouse and human TM do not indicate that the molecular mechanisms that modulate TM outflow resistance differ markedly between these two species. ^{41 42} The concept that myocilin does not act directly on TM aqueous outflow resistance is also supported by data obtained from the analysis of Myoc ^−/− mice with a targeted disruption of the Myoc gene, which have normal IOPs and no overt ocular phenotype. ⁴³  

Our in vivo results differ from in vitro findings of Fautsch et al., ²⁵ who observed a significant increase in perfusion pressure and a reduction in TM facility after perfusion of cultured human anterior segments with recombinant myocilin. They used a bacterial expression system to produce recombinant myocilin, which implies that normal eukaryotic post-translational modifications were not performed. Such modifications are relevant for myocilin which is glycosylated and forms larger complexes due to disulfide bond formation between cysteine amino acids. ^{22 44 45 46 47} Glycosylation is usually absent in recombinant proteins grown in bacteria, as prokaryotic cells are not capable of attaching carbohydrates to protein in the manner of eukaryotic cells, because bacteria lack the cell organelles (ER and Golgi apparatus) that are essential for this post-translational modification. ^{48 49 50} In addition, correct disulfide bond formation of myocilin would probably not occur in bacteria. In contrast, transgenic myocilin in the AH of βB1-crystallin-MYOC mice formed larger aggregates under nonreducing conditions, indicative of disulfide bond formation. In reducing conditions, we observed a doublet of 55 to 57 kDa for transgenic myocilin in βB1-crystallin-MYOC AH that comigrated with a similar doublet in human AH. Doublet formation of secreted myocilin has been observed previously in human AH and TM cell culture supernatant, and is caused by N-glycosylation. ^{10 44 45} Thus, the disparate outcomes between our study and that of Fautsch et al. ²⁵ could be due to differences arising from the absence of critical post-translational modifications in recombinant bacterial myocilin. Another possibility is that in both experiments different amounts of myocilin were available at those sites of the TM that are responsible for the formation of outflow resistance. Fautsch et al. ²⁵ added 25 μg of partially purified bacterial lysate (containing 25%–30% of myocilin) via an anterior chamber exchange to perfused human anterior eye segments, ²⁵ whereas, in our study, the concentration of transgenic myocilin in the AH of βB1-crystallin-MYOC was ∼0.2 ng/μL. At the present time, we do not have sufficient data for a meaningful comparison of the amounts of myocilin that were used for both experiments. It has been reported that in the mouse eye, aqueous outflow at normal IOP is difficult to measure accurately, ⁵¹ because the volume of aqueous outflow at normal IOP is small. In addition, no data are available on the effective filtration area in the mouse TM, which is obviously a magnitude smaller than that of the human TM. 

Although our findings indicate that myocilin does not directly act on TM outflow resistance, questions remain about the TM function of myocilin and the reasons that the TM expresses large amounts of myocilin. Recent in vitro data provide evidence for a de-adhesive activity of myocilin in TM cells ^{52 53} similar to that observed with matricellular proteins such as secreted protein, acidic and rich in cysteine (SPARC) or thrombospondin-1, which are also expressed in the TM. ^{54 55} Matricellular proteins are defined as secreted macromolecules with a predominately regulatory function. ^{56 57} Matricellular functions of thrombospondin-1 and SPARC appear to require receptor binding, which may also be true of a matricellular function of myocilin. In a tissue with high myocilin expression, such as the TM, additional amounts of myocilin may cause no additional effects, if putative myocilin receptors are already saturated. In Myoc ^−/− mice, SPARC and/or thrombospondin-1 may compensate for the loss of myocilin. 

It is interesting to note that transgenic myocilin was readily secreted from lenses of adult βB1-crystallin-MYOC mice and was undetectable by Western blot analysis and immunohistochemistry in the cytoplasm of lens fibers, where the chicken βB1-crystallin promoter is most active. ^{35 36} Although myocilin contains a signal sequence and has been found in AH and cell culture supernatant, ^{10 18 19 20 22 24} several investigators have discussed an intracellular role based on findings that indicated an association of myocilin with cytoskeletal elements and mitochondria of cultured cells. ^{58 59 60} Our results appear to argue against such an intracellular role of myocilin in vivo. Another aspect of myocilin secretion from transgenic lenses is that the lens capsule formed no significant barrier against the diffusion of secreted myocilin into the AH. The lens capsule is a specialized basement membrane, and critical components in the mouse capsule are collagen type IV, heparin sulfate proteoglycans, entactin-1/nidogen-1, and laminin. ^{61 62 63} The passage of myocilin through the capsule and the lack of immunostaining for myocilin in the capsule both indicate that any interactions between myocilin and lens basement membrane components are minimal. This finding correlates with observations of Filla et al., ⁶⁴ who did not observe myocilin binding to laminin and type IV collagen in solid phase binding assays, and with those of Ueda et al., ⁶⁵ who did not detect myocilin immunostaining in TM basal lamina in situ. 

In interpreting our results, it is important to keep in mind that so far, we have restricted our analysis to young adult mice that were not older than 12 weeks. In humans, there are pronounced age-related changes in the TM that result in a loss of TM cells ^{66 67} and an increase in extracellular matrix compounds. ^{68 69} Such changes might contribute to the pathogenesis of POAG. If increasing amounts of myocilin are a causative factor for the development of POAG, they may be only effective in eyes where additional age-related changes are already present. An important task for the future will be to study age-related changes in the TM of mouse eyes and any effects of myocilin on those changes. 

Our results in βB1-crystallin-Tyr437HisMYOC transgenic mice provide the first in vivo evidence that myocilin modified by a glaucoma-causing mutation is not secreted, but retained in the cell. Our data strongly indicate that the rER is the cellular compartment where mutated myocilin accumulates, whereas we observed no ultrastructural evidence for an accumulation of mutated myocilin in membrane-free, cytoplasmic inclusion bodies or aggresomes that may form when protein aggregation is disturbed during age or disease. This finding corroborates in vitro observations of others, who observed comparable findings in cultured human TM cells that had been modified to express mutated forms of myocilin. ^{10 11 12 45} A critical function of the rER is to provide quality control for correct folding of secretory and membrane proteins through the association of rER chaperones with folding polypeptide chains. ⁷⁰ Misfolded proteins are usually recognized by rER control systems, transported to the cytosol, and degraded by ubiquination and proteasomal degradation. ⁷⁰ If this mechanism fails, mutant or misfolded protein can aggregate in the rER, congest the secretory pathway, and finally evoke rER dysfunction and a cellular unfolded protein response that may ultimately lead to cell death. ^{71 72} Indeed, accumulation of misfolded protein is thought to cause cell death in several inherited neurodegenerative diseases that are—similar to POAG associated with myocilin mutations—dominant, delayed onset disorders. ^{73 74 75 76} Evidence for misfolding of mutated myocilin in TM cells in vitro has recently been observed by Liu and Vollrath, ¹¹ who also observed that the accumulation of mutant myocilin in the rER of TM cells resulted in abnormal cell morphology and cell killing. Clearly, such a scenario happens in vivo in lenses of βB1-crystallin-Tyr437HisMYOC transgenic mice, in which severe cataracts develop and causing the lenses to lose their functional capability and structural integrity. Cell death in βB1-crystallin-Tyr437HisMYOC mice was not due to apoptosis, as TUNEL-labeling of lenses was consistently negative. One should keep in mind, however, that TUNEL-labeling requires the presence of nuclear DNA, and that in adult lenses, nuclei are only present in the cells of the anterior epithelium and the bow region at the lens equator. The lens fibers that were affected most by mutated myocilin in our experiment lose their nuclei during normal lens fiber differentiation. 

It is reasonable to assume that the death of lens fibers due to an unfolded protein response was augmented by the high expression of mutated myocilin driven by the strong βB1-crystallin promoter ³⁵ and the high copy number of the transgene. In light of the comparable high expression of myocilin mRNA in the TM in situ, ^{13 16 17} it is tempting to speculate that TM cell death is the primary pathogenic event that occurs in those patients who suffer from POAG caused by mutations in myocilin. This hypothesis is supported by clinical data, as affected patients suffer from no obvious disease other than POAG, ^{26 77 78} although myocilin is expressed at lower levels in multiple tissues outside the eye, ^{13 22 79 80} and appears to play a structural role for the myelin sheath of peripheral nerves, ¹⁵ and for podocytes in the kidney. ³²  

The direct molecular mechanism on how TM cell death would lead to an increase in aqueous outflow resistance is unclear, but it is more than likely that a severe impairment of TM homeostasis caused by TM cell death should critically impair the main function of the TM. There is the possibility that mutated myocilin released from dying TM cells interacts with the extracellular matrix in the TM and directly causes or contributes to the increase in outflow resistance. Indeed, we observed small amounts of mutated myocilin in the AH of βB1-crystallin-Tyr437HisMYOC mice, which were probably released from ruptured lens fibers. This finding correlated with a significant increase in IOP that might be caused by mutated myocilin interacting with the TM outflow pathways. Nevertheless, it is far more likely that the increase in IOP in βB1-crystallin-Tyr437HisMYOC mice is not caused by mutated myocilin, but rather by the release of structural lens proteins, a scenario that is well known to cause lens-induced or phagolytic glaucoma in human patients. ⁸¹ A more direct approach targeting the TM is needed to assess putative functional roles of mutated myocilin released from dying TM cells. 

Clearly, any therapeutic intervention resulting in decreased TM expression of mutated myocilin would have the distinct potential of delaying or preventing the onset of POAG in affected patients. 

Note Added in Proof

Mice, which have been genetically modified to overexpress wild-type myocilin by an approach different from that reported in this study have been described recently (Gould DB, Miceli-Libby L, Savinova OV, et al. Genetically increasing myoc expression supports a necessary pathologic role of abnormal proteins in glaucoma. Mol Cell Biol. 2004;24:9019–9025). Similar to our results, these animals do not show significant changes in intraocular pressure. 

 Part of this work was presented at the XVI Meeting of the International Congress of Eye Research (ICER) in Sydney, Australia, September 2004.

 Supported by Deutsche Forschungsgemeinschaft Grant Sonderforschungsbereich 539, TP BI.4.

 Submitted for publication August 16, 2004; revised September 29, 2004; accepted October 9, 2004.

 Disclosure: M. Zillig, None; A. Wurm, None; F.J. Grehn, None; P. Russell, None; E.R. Tamm, 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.

 

Generation of βB1-crystallin-MYOC transgenic mice. (A) Schematic drawing of the transgene. (B) Northern blot analysis for transgenic mRNA in lenses from lines 9-16, 9-8, 9-3, and wild-type (WT) littermates (P1). Five micrograms total RNA was loaded per lane. (C) Northern blot analysis for transgenic mRNA in lenses and eyes without lens from transgenic line 9-8 and WT animals (P1). RNA from 293 EBNA cells and 293 EBNA MYOC cells, which express myocilin, was loaded as a control. 28S ribosomal RNA was stained with methylene blue. (D) In situ hybridization for transgenic mRNA at E13.5. In the lens (Le) of transgenic animals (TG), myocilin mRNA was detected in lens fibers, but was absent in lens epithelium or other ocular tissues. Lenses of wild-type (WT) littermates were essentially negative. Re, retina. Scale bars, 88 μm.

Figure 2.

 

Secreted transgenic myocilin in βB1-crystallin-MYOC transgenic mice. (A) Western blot analysis for transgenic myocilin in lenses from lines 9-3, 9-8, and 9-16 and wild-type (WT) littermates at P1. Different amounts of purified recombinant myocilin (rMyoc) were loaded for comparison. (B) Western blot for transgenic myocilin in lens and AH of 9-16 animals and wild-type littermates (WT) at P21. (C) Western dot blot for transgenic myocilin in AH samples from four transgenic animals (T1–T4), four wild-type littermates (C1–C4), and six human donors (H1–H6). Total protein was visualized by fluorescent protein gel staining (SYPRO Ruby; Molecular Probes). Different amounts of BSA were loaded for comparison. (D) Western blot for transgenic myocilin in the AH from one human donor (20 μL), transgenic 9-16 animals (TG) and wild-type littermates (WT) at P21. Five microliters AH was obtained from one animal, and 17 μL was pooled from the eyes of four animals. Different amounts of purified recombinant myocilin (rMyoc) were loaded for comparison. Eight percent to 15% gradient (A, D) or 12% (B) SDS-polyacrylamide gels were used.

Figure 3.

 

Immunostaining for myocilin in the eyes of βB1-crystallin-MYOC transgenic animals (TG; B, D, F) and wild-type littermates (WT; A, C, E) at P14. Rabbit antibodies against myocilin were used. (A, B) Positive immunoreactivity for myocilin was detected in a thin layer (arrows) on the lens (Le) surface of transgenic animals (TG), but not in wild-type (WT) littermates. (C, D) Keratocytes and corneal endothelial cells were positively labeled in the cornea (Co) of both wild-type (WT) and transgenic (TG) animals. In addition, in transgenic animals, the inner surface of the cornea was covered with a thin layer of immunoreactive material (arrows). (E, F) Cells of the ciliary body (CB) and cells of the chamber angle were positively stained for myocilin in both transgenic (TG) and wild-type animals (WT). In addition, in transgenic animals, a homogenous mass with strong immunoreactivity for myocilin was observed in the chamber angle and covered the inner parts of the TM (arrow). Scale bars: (A, B) 16 μm; (C–F) 10 μm.

Figure 4.

 

Light (A–D) and electron (E, F) microscopy of eyes from βB1-crystallin-MYOC transgenic (A, C–F) and wild-type (B) animals at 12 weeks (wks) of age. (A, B) In both transgenic (A) and wild-type (B) animals, the structure of ciliary body (CB), iris, anterior chamber (AC), TM, and Schlemm’s canal (arrow) were essentially normal. (C) Higher magnification of the TM in (A). The extracellular spaces in the TM of transgenic animals were optically empty (star). (D) No structural changes were noted in lens fibers in the bow region of transgenic lenses, where transcription of the transgene occurs. (E, F) Electron microscopy of the TM in transgenic animals. (E, box) Site shown at higher magnification in (F). Trabecular cells were structurally intact. In both the corneoscleral parts of the TM (E, star), and the outer juxtacanalicular part close to the endothelial lining of Schlemm’s canal (F, star), the extracellular spaces were largely empty (E, F). In the juxtacanalicular part, some extracellular fine fibrillar material was observed (arrows), the amount of which was not different from that in control animals. Scale bars: (A, B) 40 μm; (C) 10 μm; (D) 16 μm; (E) 1.9 μm; (F) 0.42 μm.

Figure 5.

 

Intraocular pressure (IOP) readings in 9-16 transgenic animals and wild-type littermates (WT) at 12 weeks (wks) of age. There was essentially no difference in IOP between both groups of animals.

Figure 6.

 

Generation of βB1-crystallin-Tyr437HisMYOC transgenic mice. (A) Schematic drawing of the transgene. (B) Northern blot analysis for transgenic mRNA in lenses of transgenic lines 8-5, 8-6, 8-13 and wild-type (WT) littermates (P1). (C) Northern blot analysis for transgenic mRNA in lenses and eyes without lens of 8-5 transgenic animals and wild-type (WT) control animals (P1). Five micrograms total RNA was loaded per lane. RNA from 293 EBNA cells and 293 EBNA MYOC cells that express myocilin was loaded as the control. (D) Western blot analysis (12% SDS-polyacrylamide gel) for transgenic mutated myocilin in lens and AH of 8-6 transgenic animals (TG) and wild-type littermates (WT) at P21. A strong signal for mutated myocilin was detected in the lens, whereas only a weak band was observed in AH.

Figure 7.

 

Light microscopy (A, B), electron microscopy (C–F), and immunostaining for myocilin (G, H) in lenses of βB1-crystallin-Tyr437HisMYOC transgenic mice (B, D, F, H) and wild-type littermates (A, C, E, G) at P21. (A, B) Multiple vesicles (arrows) were noted seen in lens fibers at the bow region of transgenic animals (B), but not in wild-type littermates (A). (C, D) By electron microscopy, the vesicles (D, arrow) were localized to the cytoplasm surrounding the nucleus (N) of lens fibers and were filled with electron-dense granular material (D). In wild-type littermates (C), the same perinuclear area of lens fibers contained cisterns of rER, which appeared to be of normal size (C, arrow). (E, F) Higher magnification of perinuclear vesicles in transgenic animals (F) showed that the vesicles were surrounded by a membrane that contained ribosomes (F, arrows) and confirmed their origin from rER cisterns, which were of normal size in wild-type littermates (E, arrows). (G, H) Immunocytochemistry with rabbit antibodies specific for myocilin showed strong positive immunoreactivity in lens fiber vesicles (arrows) of transgenic animals, and confirmed that the vesicles were caused by an accumulation of Tyr437His mutated myocilin in the rER. Scale bars: (A, B) 16 μm; (C, D) 690 nm; (E, F) 166 nm; (G) 6 μm; (H) 4 μm.

Figure 8.

 

Cataract developed in lenses of βB1-crystallin-Tyr437HisMYOC transgenic mice. (A, B) Lenses of transgenic mice expressing Tyr437His mutated myocilin lost transparency because of nuclear cataracts (A), whereas lenses of wild-type littermates (B) were transparent. (C) Light microscopy of the posterior pole of a transgenic lens (Le) with cataract that was ruptured at its posterior pole (arrows). Age of animals, 6 weeks (wks). Re, retina. Scale bars (C) 12 μm.

Figure 9.

 

IOP readings in 8-6 transgenic animals and wild-type littermates (WT) at 12 weeks (wks) of age. Mean IOP was significantly higher ( , P < 0.004) in animals overexpressing mutated myocilin than in wild-type littermates.

The authors thank Kathrin Baier, Karin Göhler, Antonia Kellenberger, Jasmin Onderka, and Nadine Petersen for excellent technical help, Marco Gösswein for expert processing of the electron micrographs, and Eric Wawrousek of the NEI Transgenic Facility for his invaluable help.