Optical dysfunction of the crystalline lens in aquaporin-0-deficient mice

ALAN SHIELS1,2, STEVEN BASSNETT1,3, KULANDAIAPPA VARADARAJ4, RICHARD MATHIAS4, KRISTIN AL-GHOUL5,6, JER KUSZAK6,7, DORIT DONOVIEL8, STAN LILLEBERG8, GLENN FRIEDRICH8 and BRIAN ZAMBROWICZ8

1 Departments of Ophthalmology and Visual Sciences
2 Genetics
3 Cell Biology Washington University School of Medicine, St. Louis, Missouri 63110
4 Department of Physiology, State University of New York, Stony Brook, New York 11794
5 Departments of Anatomy
6 Pathology
7 Ophthalmology, Rush-Presbyterian-St. Luke’s Medical Center, Chicago, Illinois 60612
8 Lexicon Genetics, The Woodlands, Texas 77381


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Aquaporin-0 (AQP0), a water transport channel protein, is the major intrinsic protein (MIP) of lens fiber cell plasma membranes. Mice deficient in the gene for AQP0 (Aqp0, Mip) were generated from a library of gene trap embryo stem cells. Sequence analysis showed that the gene trap vector had inserted into the first exon of Aqp0, causing a null mutation as verified by RNA blotting and immunochemistry. At 3 wk of age (postnatal day 21), lenses from null mice (Aqp0-/-) contained polymorphic opacities, whereas lenses from heterozygous mice (Aqp0+/-) were transparent and did not develop frank opacities until ~24 wk of age. Osmotic water permeability values for Aqp0+/- and Aqp0-/- lenses were reduced to ~46% and ~20% of wild-type values, respectively, and the focusing power of Aqp0+/- lenses was significantly lower than that of wild type. These findings show that heterozygous loss of AQP0 is sufficient to trigger cataractogenesis in mice and suggest that this MIP is required for optimal focusing of the crystalline lens.

cataract; gene trap; mouse; water channel


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
THE CRYSTALLINE LENS is an ocular component of high transparency and refractive index that is responsible for variable focusing of light rays onto the photosensitive retina. In vertebrates, the lens develops from a patch of head ectoderm that invaginates to form a hollow epithelial vesicle within the embryonic eye (24). Postmitotic cells at the posterior of the vesicle elongate to fill the cavity with primary fibers forming the embryonic lens. The anterior monolayer of epithelial cells retains lifelong mitotic capacity and, subsequently, the lens continues to grow by laying down concentric shells of secondary fibers that elongate from epithelial cells within the germative zone at the lens equator. The most abundant cell surface marker of this terminal differentiation process is a hydrophobic, transmembrane polypeptide of molecular mass ~26 kDa (13) encoded by the gene for major intrinsic protein (MIP). Remarkably, MIP is the charter member of a sequence-related gene family encoding over 150 channel proteins that facilitate the transport of water (aquaporins) and glycerol (aquaglyceroporins) across the plasma membranes of microbial, plant, and animal cells (12). Significant water (but not glycerol) transport has been attributed to MIP in lens fiber cell membranes (32), and this pH- and calcium-sensitive water channel (23) is commonly referred to as aquaporin-0 (AQP0).

Of the nine other aquaporin genes identified in mammals, at least six (AQP1–4, AQP6, and AQP7) are expressed in the kidney (3), an organ largely devoted to high rates of active fluid transport. Aquaporins are also widely expressed in nonrenal tissues (3), including red blood cells (AQP1) lung (AQP5), gastrointestinal tract (AQP6), liver (AQP9), brain (AQP4), skeletal muscle (AQP4), secretory glands (AQP5), and eye (AQP0, AQP1, and AQP3–5). Consistent with the complex tissue distribution of water channels, the first reports of aquaporin gene mutations have displayed diverse phenotypes. In humans, recessive mutations in AQP1 were associated with loss of the Colton blood group in otherwise normal individuals (25), whereas mutations in AQP2 resulted in recessive (7) and dominant (16) forms of nephrogenic diabetes insipidus, an inability to concentrate urine. Dominant mutations in the gene for AQP0 have been associated with lens opacities, or cataracts, in humans (11) and mice (26, 28, 29). In addition, a number of aquaporin deletion mice have been generated by targeted gene disruption through homologous recombination. Such mice lacking AQP1 (20), AQP2 (34), AQP3 (19), or AQP4 (4) suffered mild (AQP4) to severe (AQP2) polyuria and dehydration following water deprivation, whereas AQP5 deletion resulted in impaired saliva secretion (18). Significantly, heterozygous loss of these water channels was not sufficient to produce an obvious phenotype. To determine the impact of AQP0 deficiency on lens function, we have generated AQP0 deletion mice using a gene trapping strategy (35) that allows random targeting of genes in embryonic stem (ES) cells.


    METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

Gene trap mice.
Mouse ES cells, derived from a 129S5 strain background, were transduced with a recombinant retroviral construct for gene trapping (VICTR20) followed by puromycin selection of resistant clones to produce the OmniBank library of retrovirally mutagenized ES cell lines (35). VICTR20 (Fig. 1A) contains a sequence acquisition component consisting of the murine phosphoglycerate kinase-1 promoter, which is active in ES cells, fused to the puromycin N-acetyltransferase gene coding sequence that lacks a polyadenylation signal but is followed by a synthetic consensus splice donor sequence (PGKpuroSD). Puromycin-resistant ES cells resulted from splicing of PGKpuroSD to downstream exons of a trapped gene and addition of a polyadenylation signal producing a fusion transcript containing puro sequences at the 5' end and sequences from the trapped gene at the 3' end. Total RNA was extracted from puromycin-resistant ES cells with RNAzol (Tel-Test), then reverse transcribed into cDNA and screened for the presence of the puro-gene trap transcript using 3' rapid amplification of cDNA ends (3'-RACE) and direct sequencing using the OBS sequencing primer (35). One mutant ES cell line, identified by an OmniBank sequence tag (OST1231), matched the known Aqp0 sequence beginning in the second exon (http://omnibank.lexgen.com). Chimeric mice were derived from injections of OST1231 ES cells into C57BL/6J-albino host blastocysts, and heterozygous mice from breeding high contribution male chimeras to wild-type C57BL/6J-albino females. All intercrosses among heterozygous and homozygous animals were in a hybrid 129S5/C57BL/6J-albino background. Mice were barrier-housed in 12:12-h light/dark cycles with free access to food and water in accordance with the animal studies regulations at Washington University. Mice were killed by CO2 asphyxiation followed by cervical dislocation or decapitation. Eyes were enucleated from age-matched littermates, and lenses were dissected in PBS at 37°C then photographed under a dissecting microscope (Zeiss, Stemi 2000). Prior to wet-weight measurements, individual lenses were gently rolled on PBS-soaked tissue paper to remove remnants of ciliary-body epithelium from the equatorial region. Preweighed lenses were then dried to constant weight in a vacuum oven at 60°C as described previously (1), then reweighed.



View larger version (44K):
[in this window]
[in a new window]
 
Fig. 1. Generation of Aqp0-deficient mice by gene trapping. A: schematic diagram of the gene trap vector (VICTR20) site of insertion into Aqp0. Arrows indicate the orientation and approximate location of PCR primers listed in Table 1. B: sequence of the novel junction between the 3'-LTR of the vector (bold lower case) and exon 1 of Aqp0. Asterisk indicates an in-frame translation stop codon predicted to truncate the protein at valine-56. C: PCR-based genotyping of wild-type (+/+), heterozygous (+/-), and null (-/-) mice showing the wild-type allele (277 bp) using primers Mip.4 and Mip.5 and the mutant allele (788 bp) using primers Mip.4 and Rv.3 (A and Table 1). D: RNA blot analysis of Aqp0 transcripts in eye (E), heart (H), brain (B), lung (L), and muscle (M) tissues from wild-type (+/+), heterozygous (+/-), and null (-/-) mice. E: immunoblot analysis of AQP0 in lens proteins from wild-type (+/+), heterozygous (+/-), and null (-/-) mice.

 
PCR and sequence analyses.
The position of VICTR20 insertion within the Aqp0 gene was determined by PCR amplification of DNA fragments flanking the long terminal repeats (LTRs). Genomic DNA from tail biopsies of heterozygotes (1.0 µg) was diluted to 10 µl with water, denatured at 96°C for 3 min, and then held at 80°C. PCR master mix (40 µl) was added, and amplification was initiated. The final concentration of each PCR component was as follows: 1x PCR buffer (50 mM Tris·HCl, 15 mM ammonium sulfate [pH 9.3]), 2.5 mM MgCl2, 0.1% Tween 20, 500 µM dNTPs, 5% DMSO, 0.5 µM VICTR20-specific primer (19), 0.5 µM Aqp0-specific primer (Table 1), and 0.05 U/µl of AccuTaq LA DNA polymerase (Sigma). Amplification cycle conditions were as follows: denaturation at 96°C for 40 s, annealing and extension at 68°C for 6 min repeated for 30 cycles. A final extension time of 12 min was also included. Nested PCR was conducted with 3.0 µl of a 100-fold dilution of the primary PCR sample using nested VICTR20- and Aqp0-specific primers (Table 1). The nested reaction composition and cycle parameters were the same as for the primary PCR. DNA fragments obtained in these amplification reactions were sequenced directly using the OBS sequencing primer (35). The orientation and sequence of VICTR20-specific and Aqp0-specific primers used for PCR amplification are shown in Fig. 1A and Table 1, respectively. For sequences downstream (3') of the gene trap site, outer primers Po and Mip-R1, then nested primers Pi and Mip-R2 were used. For sequences upstream (5') of the gene trap site, outer primers Zo and Mip-F1, then nested primers Zi and Mip-F2 were used. Mice were genotyped by PCR amplification of tail DNA with a combination of three primers, i.e., Mip.4, Mip.5, and Rv.3 (Table 1). Mip.4 and Rv.3 amplify a band (788 bp) specific for the mutant allele, and Mip.4 and Mip.5 amplify a band (277 bp) specific for the wild-type allele.


View this table:
[in this window]
[in a new window]
 
Table 1. PCR primers used to characterize the Aqp0 gene trap

 
ß-Galactosidase activity.
Embryos from wild-type females crossed with heterozygous males were collected at 12–13 days post-coitus (dpc), fixed in 2% paraformaldehyde/0.2% glutaraldehyde (wt/vol in PBS), then stained with 5-bromo-4-chloro-3-indolyl-{alpha}-D-galactopyranoside (X-gal), essentially as described previously (21), and photographed under a phase-contrast microscope. In addition, fixed lenses were sectioned as described below and stained using the GAL-S ß-galactosidase reporter gene staining kit (Sigma).

In situ hybridization.
Whole mount in situ hybridizations were performed essentially as detailed previously (5). Briefly, embryos (~13 dpc) were fixed (2 h at 4°C) in paraformaldehyde (4% wt/vol in PBS), permeabilized in 0.1% Tween/PBS (PBST), dehydrated in methanol, and stored (-20°C). After rehydration, fixed embryos were digested with proteinase K (20 µg/ml in PBST) and refixed in 4% paraformaldehyde/0.2% glutaraldehyde. Following prehybridization (50% formamide, 1 h at 63°C), embryos were hybridized (16 h, 63°C) with a digoxigenin-UTP-labeled cRNA probe, derived from a 470-bp fragment (nt 293–763) of mouse Aqp0 cDNA (GenBank accession no. U27502), as described by the kit manufacturer (Roche Biochemicals). Following RNase treatment (100 µg/ml, 37°C), embryos were incubated (16 h, 4°C) with digoxigenin antibody coupled to alkaline phosphatase, then stained (16 h, dark) with nitroblue tetrazolium (NBT) and 5-bromo-4-chloro-3-indolyl phosphate (BCIP) in the presence of levamisole (2 mM).

Northern blot analysis.
Total RNA was extracted from multiple tissues from 8-wk-old littermates, representing each of the three genotypes, using RNAzol (Tel-Test). RNA (~10 µg) was separated on 1% agarose-formaldehyde denaturing gels, transferred to nylon, and probed at high stringency, essentially as described previously (27), with the 277-bp PCR product amplified from genomic DNA by primers Mip.4 and Mip.5 (Table 1) in the presence of [{alpha}-32P]dCTP.

Immunoblot analysis.
Lenses were homogenized in PBS containing a protease inhibitor cocktail (Complete-Mini tablets, Roche Biochemicals), and protein concentration was quantified by NanoOrange (Molecular Probes) using a fluorometer (Bio-Rad). Proteins were separated on 12% SDS-PAGE gels (Bio-Rad) then immunoblotted with AQP0 antibody (Chemicon) using a Mini Protean II system (Bio-Rad) and enhanced chemiluminescence (ECL) detection reagents (APBiotech) as described previously (28).

Immunofluorescence microscopy.
Embryo eyes at 15.5 dpc were fixed in 4% paraformaldehyde, embedded in agar then sectioned (50–100 µm), using a Vibratome as described previously (28). Sections were permeabilized with Triton X-100, incubated with either polyclonal AQP0 antibody (Chemicon) followed by fluorescein-conjugated goat anti-rabbit immunoglobulin G (Sigma) or Texas Red-phalloidin (Molecular Probes), an F-actin-specific probe, as described previously (2). Stained sections were then viewed on a laser scanning confocal microscope (Zeiss model LSM410).

Water permeability measurements.
Lens fiber cell vesicles were incubated in hypertonic saline (450 mm), and the rate of shrinkage was imaged by digital video microscopy as described previously (32). The presence or absence of AQP0 in vesicle preparations was confirmed by immunofluorescence as described previously (32).

Optical quality analysis.
Within 5 min of death, excised lenses (~2.5 mm diameter) from wild-type and Aqp0+/- mice were assessed by a Scantox In Vitro Assay System generously provided by Harvard Apparatus (Holliston, MA). This system consists of a collimated laser source that projects its light onto a plain mirror mounted at 45° on a carriage assembly. The mirror reflects the laser beam directly up through the lens under examination. Each lens was suspended on a beveled washer designed to support its equatorial rim within a specially designed two-chambered cell made of glass and silicon rubber. Both anterior and posterior lens surfaces were bathed in prewarmed (37°C) culture medium (25 ml) consisting of M199 with Earle’s salts and 8% fetal bovine serum (Invitrogen). Periodic measurements of medium osmolarity and pH indicated little or no change over the duration of the experiment. The mirror carriage was connected via a drive screw to a positioning motor that moved the laser across the lens in a series of equal increments, defined by dividing the equatorial diameter of the lens by the number of steps. An initial series of eight laser scans were passed through each lens. A mechanical linkage then made it possible to rotate the lens through an angle of 90° to accomplish a second series of eight scans. Thus each lens was scanned a total of 16 times for a grand total of 288 objective measurements in this study. A digital camera captured the actual position and slope of the laser beam at each step. The captured data is used to calculate the back focal length (BFL) of each position and the difference in BFL between beams. BFL was defined as the back vertex distance (mm) measuring the laser beam from the rear surface of the lens to the focal point. Repeated measurements of BFL indicated that instrument reproducibility was within ±0.32% of focal length. Changes in this distance with beam position (gross focus) were a result of longitudinal spherical aberration. In addition, morphological irregularities of the lens can misdirect the laser beam, causing a more scattered appearance of the focal point. Therefore, we also measured BFL variability (BFLV), to quantify the sharpness of focus, defined as the average standard error of the means of the BFL of all lenses in each group. Statistical calculations to determine whether significant differences were present between the average BFL and BFLV were carried out using a t-test for samples of independent types. A probability value of at least P <= 0.05 was considered significant.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

Gene trapping of Aqp0.
The gene trap construct (VICTR20) was designed such that the trapping genes are transcriptionally oriented in the direction opposite to that of its retroviral vector (35). Therefore, sequences flanking the 3'-LTR of VICTR20 would be the most 5' relative to the mouse AQP0 gene (Aqp0, Mip), which is organized into four exons (Fig. 1A). The OmniBank sequence tag (OST) 1231, derived from 3'-RACE analysis of heterozygous mutant ES cells (http://omnibank.lexgen.com), began in the second exon of Aqp0, indicating that the gene trap event had occurred in either the first intron or the first exon of the gene. To confirm that the retrovirus was integrated into the 5' end of Aqp0, a genomic DNA fragment adjacent to the 3'-LTR of VICTR20 was amplified from OST1231 ES cells using nested primer pairs (Table 1) located in exon 1 (Mip-F1, Mip-F2) and in the LacZ reporter gene (Zo, Zi). Sequence analysis of the resulting PCR product identified a novel junction between the 3'-LTR and the first exon of Aqp0 at a position located 168 nucleotides downstream from the translation start codon (Fig. 1B). The LTR fusion introduced an in-frame translation stop signal (tga) at codon 57 of Aqp0 essentially mimicking a nonsense mutation. The resulting chain termination at valine-56 was predicted to abolish ~79% (207 amino acids) of the AQP0 protein including both of the functionally conserved asparagine-proline-alanine (NPA) motifs (15), located at codons 68–70 and 184–186.

Following the generation of chimeric animals, a PCR strategy with three primers (Table 1) in one reaction was implemented to genotype progeny with wild-type or mutant Aqp0 alleles (Fig. 1C). Two primers (Mip.4 and Mip.5) specific for Aqp0 exon 1 amplified a 277-bp fragment from the wild-type allele. The mutant allele was detected from a combination of the 5'-exon 1 primer (Mip.4) and a retrovirus-specific primer (Rv.3) resulting in a 788-bp fragment. The official allele symbol for this gene trap mutation is MipGt(VICTR20)8Lex.

To confirm that the insertion of VICTR20 generated a null allele, we compared the level of Aqp0 transcripts and translation products in wild-type and mutant progeny. Northern blot analysis (Fig. 1D) detected characteristic Aqp0 transcripts of ~2.2 kb and ~1.2 kb (27) specifically in eye-derived RNA from wild-type mice. Aqp0 transcript levels in heterozygous mutants were reduced relative to wild-type levels, and no Aqp0 transcripts were detected in homozygous mutants. Similarly, immunoblot analysis of lens homogenates (Fig. 1E) detected reduced levels Aqp0 antigen (Mr ~26 kDa) in heterozygous mutants, but no such protein was detected in homozygous mutants.

Embryonic expression of the LacZ reporter gene and Aqp0.
In the mutant Aqp0 allele, splicing was predicted to occur between a cryptic splice-donor site in the 3'-LTR and the splice-acceptor site in the gene trap vector (Fig. 1A). Consequently, the Aqp0 promoter was positioned to drive transcription of the adjacent ß-geo reporter gene, which codes for a bifunctional translational fusion between ß-galactosidase and neomycin phosphotransferase (35). Thus, to monitor Aqp0 promoter activity in situ we used X-gal staining of whole mount embryos at 12–13 dpc (Fig. 2). ß-Galactosidase activity was detected in the lens and periocular tissue, within ~1 h of incubation in the X-gal substrate (Fig. 2A). On longer incubation, ß-galactosidase activity was also visible in the rib cage and nasal regions (Fig. 2A). Within the lens, ß-galactosidase activity was restricted to fiber cells and absent from anterior and equatorial epithelial cells (Fig. 2C).



View larger version (76K):
[in this window]
[in a new window]
 
Fig. 2. Expression pattern of the ß-galactosidase reporter gene in Aqp0+/- mouse embryos [12–13 days post-coitus (dpc)]. A: whole mount heterozygous (+/-) embryo showing LacZ staining (blue) of lens, periocular, snout, and rib regions. No such ß-galactosidase activity was detected in the age-matched wild-type (+/+) embryo B: the eye region is outlined by the retinal pigmented epithelium (brown). C: sagittal section (polar plane) through the heterozygous (+/-) lens (anterior pole up) showing that LacZ staining was confined to fiber cells extending along the anterior/posterior (polar) axis and absent from the anterior epithelial cell layer. Remnants of the embryonic lens vasculature can be observed at the posterior pole. D: whole mount in situ hybridization of the wild-type (+/+) embryo head region showing that Aqp0 transcripts are confined to the lens and absent from the periocular and snout regions.

 
To correlate transcription of Aqp0 with that of the ß-galactosidase reporter gene, we used in situ hybridization analysis of whole mount embryos at 13 dpc. Figure 2D shows that the Aqp0 probe hybridized strongly in the lens but not in the periocular or nasal regions of the wild-type embryo. Similarly, Aqp0 hybridization was not detected in the wild-type rib regions that were ß-galactosidase positive in Aqp0-/- embryos (data not shown). Although we cannot exclude the possibility that the Aqp0 hybridization probe did not adequately penetrate the embryo to stain the ribs, the lack of superficial periocular hybridization indicated that the non-lens ß-galactosidase activity in Aqp0+/- embryos (Fig. 2A) did not recapitulate expression of the endogenous gene.

To further evaluate Aqp0 translation in situ, we performed immunofluorescence confocal microscopy on embryonic (13–15 dpc) lens sections with AQP0 antibody. Specific and intense immunostaining of fiber cell plasma membranes was detected in the wild-type lens (Fig. 3A). A similar pattern of fiber cell immunostaining was detected in Aqp0+/- lenses (Fig. 3B), reflecting the sheer abundance of AQP0 in the lens; however, the uniformity and ordered arrangement of fiber cells along the polar axis of the lens had been compromised. No immunostaining of fiber cells was detected in Aqp0-/- lenses (Fig. 3C), consistent with the AQP0 immunoblot analysis of these lenses (Fig. 1E). However, phalloidin staining of the actin cytoskeleton confirmed that irregular, fiber-like cells had elongated toward the poles of the Aqp0-/- lens (Fig. 3D).



View larger version (75K):
[in this window]
[in a new window]
 
Fig. 3. Immunofluorescence confocal microscopy of the Aqp0-deficient mouse lens (15.5 dpc). A: sagittal section (polar plane) of a wild-type (+/+) lens (anterior pole up) showing intense immunostaining of fiber cell membranes with AQP0 antibody (red channel). B: heterozygous (+/-) lens showing an AQP immunostaining pattern similar to that of wild type. C: null (-/-) lens showing complete absence of AQP0 immunostaining. D: null (-/-) lens showing phalloidin staining of actin filaments (green channel). Cell nuclei in AC have been stained with propidium iodide (blue channel) to display their characteristic bow-shaped pattern. Scale bars, 100 µm.

 
Gross pathology of the Aqp0-deficient lens.
The hybrid 129S5/C57BL/6J-albino background segregated wild-type, heterozygous, and homozygous progeny at Mendelian frequencies in the F2 generation. There was no obvious sign of reduced eye size (microphthalmia), and lens opacities were difficult to detect in the pigmented eyes of live agouti and black Aqp0-/- progeny without the aid of an ophthalmoscope. However, lens opacities were evident in albino Aqp0-/- progeny when the eyes opened around postnatal day 14 (P14) (data not shown), consistent with the congenital onset of cataracts observed in mice homozygous for either a transposon-induced splice mutation (CatFr) or a missense (A51P) mutation (Catlop) in Aqp0 (26, 28). In contrast, Aqp0+/- mice did not develop frank lens opacities in vivo until ~6 mo of age (data not shown), considerably later than the age-of-onset observed in heterozygous Catlop (birth) and CatFr (~6 wk) mice. Such a delay in Aqp0+/- cataract onset may be related to the absence of mutant Aqp0 translation products, which accumulate in Catlop and CatFr lenses (26, 28); however, we cannot exclude the possibility that genetic background differences between the three mouse strains may also be involved. In common with the spontaneous Catlop and CatFr mutant mice, lens opacities were bilateral and penetrant in all Aqp0+/- (n > 50) and Aqp0-/- (n > 50) progeny examined.

To further evaluate Aqp0-deficient pathology, lenses were excised at 3 wk of age (P21) and photographed under warm media to prevent the formation of ambient temperature-induced ("cold") opacities. Compared with wild-type and Aqp0+/- lenses, Aqp0-/- lenses were more fragile and prone to rupture during dissection from the eye. When viewed from the anterior or posterior poles under oblique illumination, the P21 Aqp0+/- lens (Fig. 4B) displayed an abnormal transparent zone of light refraction that was not present in the wild-type lens (Fig. 4A). This optically disturbed zone resembled patterned glass and extended from the central core or nucleus into the peripheral cortex of the lens. Similar illumination of the P21 Aqp0-/- lens revealed dense polymorphic opacities surrounded by vacuolated and translucent regions of light scattering (Fig. 4C).



View larger version (49K):
[in this window]
[in a new window]
 
Fig. 4. Gross pathology of the Aqp0-deficient mouse lens at 3 wk of age (postpartum day 21, P21). A: dissecting microscope image of wild-type (+/+) lens (posterior pole) under oblique illumination. B: similar view of heterozygous (+/-) lens showing the optically distorted core region. C: null (-/-) lens showing central opacity surrounded by a polymorphic region of light scattering.

 
Water content and permeability of the Aqp0-deficient lens.
Wet and dry weight measurements were used to determine the water content of age-matched (P21) lenses. Mean wet weight values (±SE) increased from 3.09 ± 0.05 mg for wild-type lenses (n = 21) to 3.36 ± 0.05 mg for Aqp0+/- lenses (n = 26) and decreased to 2.88 ± 0.16 mg for Aqp0-/- lenses (n = 14). Mean dry weight values decreased from 0.87 ± 0.03 mg for wild-type lenses to 0.82 ± 0.03 mg for Aqp0+/- lenses and 0.64 ± 0.07 mg for Aqp0-/- lenses. Wet weight values for Aqp0+/- lenses and dry weight values for Aqp0-/- lenses were significantly (P <= 0.05) different from wild type; however, wet weight and dry weight values for Aqp0-/- and Aqp0+/- lenses, respectively, were not. When lenses were adjusted for dry weight, the water content was found to increase from ~71.8% in wild-type lenses to ~75.6% in Aqp0+/- lenses and ~77.8% in Aqp0-/- lenses, consistent with the increasing severity of gross pathology (Fig. 4).

To determine the osmotic water permeability (Pf) of Aqp0-deficient lenses, we measured the shrinkage rate of fiber cell membrane vesicles in response to osmotic challenge (300 mM to 450 mM saline) in vitro. For vesicle preparation, we chose age-matched lenses at 3–4 wk of age (P21P28) to obtain sufficient non-degraded cortical fiber cell membrane material for study. Figure 5 shows that the Pf values decreased from 39 ± 11 µm/s in wild-type vesicles (n = 9), to 18 ± 5 µm/s in Aqp0+/- vesicles (n = 6) and to 8 ± 3 µm/s in Aqp0-/- vesicles (n = 16), representing decreases of ~54% and ~80%, respectively. Notably, the Pf values of the Aqp0-/- lenses did not decrease to those of fiber cell membrane lipids alone (1.5 ± 0.2 µm/s, Ref. 32), suggesting that other membrane proteins and/or changes in lipid composition partially compensated for lack of AQP0-mediated water transport.



View larger version (14K):
[in this window]
[in a new window]
 
Fig. 5. Osmotic water permeability (PH2O) of lens fiber cell membrane vesicles from Aqp0-deficient mice at 3–4 wk of age (P21P28). A: wild-type (+/+) vesicles. B: heterozygous (+/-) vesicles. C: null (-/-) vesicles. Each panel shows a typical profile and gives the mean PH2O value ± SD for n vesicles. Vi(t), internal volume (cm3) of the vesicle; {tau} = time constant (s) of the change in volume determined by curve fitting.

 
Optical quality of the Aqp0-deficient lens.
To evaluate the optical quality of the Aqp0-deficient lens, we used a low-power helium-neon laser scanning monitor system to measure the average BFL, which is a measure of longitudinal spherical aberration, and the BFLV, which is a measure of focusing power as a function of internal structure (30). To obtain lenses large enough (>2 mm diameter) to permit multiple passes of the laser, we chose age-matched mice at 5–6 mo of age. Consequently, we were unable to obtain valid data for Aqp0-/- lenses, as they were almost totally opaque at this age. Figure 6 shows representative laser scan profiles from age-matched wild-type and Aqp0+/- lenses. It was readily apparent that laser beams passed through wild-type lenses came to a sharper focus than those passed through Aqp0+/- lenses. Results of statistical analysis (t-tests for independent groups of lenses) quantifying the differences between average BFL and BFLV of wild-type vs. Aqp0+/- lenses (Table 2) confirmed that optical quality (sharpness of focus) was significantly degraded in the latter.



View larger version (23K):
[in this window]
[in a new window]
 
Fig. 6. Representative laser-scan profiles of age-matched (5–6 mo) wild-type and Aqp0+/- mouse lenses. Horizontal axis indicates the back focal length (mm), and the vertical axis indicates the distance (mm) from the optical center (0.0) of the lens of the series of incident laser beams passed through a lens. A: wild-type (+/+) lens. Note all laser beams intersected at a fairly sharp focal point. B: heterozygous (+/-) lens. In contrast, these laser beams failed to intersect at a consistent focal point.

 

View this table:
[in this window]
[in a new window]
 
Table 2. Statistical analysis of average BFL and BFLV in wild-type vs. Aqp0+/- mouse lenses

 

    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Two aquaporins have been detected in the crystalline lens. AQP1 is confined to the mitotic epithelial cell layer that lines the anterior surface (14), whereas AQP0 is characteristic of the terminally differentiated fiber cells (33) that constitute the bulk of the lens mass. Although AQP1 is at least 10-fold more active as a water channel than its fiber cell counterpart (23), no spontaneous lens pathology has been associated with AQP1 deficiency in humans (25) or mice (20). In this study, we have shown that AQP0 accounts for ~80% of the water permeability of mouse lens fiber cell plasma membranes and that heterozygous loss of this MIP is sufficient to compromise lens transparency. These observations suggest that although the lens can compensate for loss of AQP1 function, there is no such redundancy for AQP0 deficiency.

The relative loss of water permeability recorded in Aqp0+/- mouse lenses was similar in magnitude to that measured in the kidney proximal tubules of Aqp1+/- mice (20). This loss of AQP0 function was further associated with a significant reduction in the focusing power of the Aqp0+/- lens compared with wild type. The precise role of water transport in the lens is unclear. One hypothesis has proposed that a microcirculation of electrolytes throughout the avascular lens may be coupled to water flow (22). Moreover, significant growth-related changes in water content have been associated with maintaining lens optical quality (8, 9). Thus, as new fibers are added at the lens equatorial cortex, they undergo a maturation process that gradually removes water and increases the concentration of the transparent cytoplasmic proteins, or crystallins. Consequently, the refractive index of the lens nucleus is significantly higher than that of the peripheral cortex. Thus, by acting as a conduit for fiber cell dehydration, it is conceivable that AQP0 may play a role in establishing the refractive index of the crystalline lens.

The sheer abundance of AQP0 in fiber cells also raises the possibility that this comparatively inefficient water channel has been recruited to share some other function(s) in the lens. Ultrastructural studies have indicated that AQP0 can form specialized contacts between mature fibers within the lens nucleus (6, 36). Furthermore, recent crystallographic studies have detected tongue-and-groove contours on the extracellular surface of reconstituted AQP0 tetramers that may facilitate fiber cell adherence in the lens (10). Normally, the uniform hexagonal shaped fiber cells, interfaced on all six sides with comparable fiber cells, are arranged into highly ordered and very closely apposed radial cell columns and growth shells (17). The elimination of extracellular space between fibers is considered an essential part of transforming the lens into a series of coaxial refractive surfaces (31). Thus it is also possible that AQP0 deficiency results in an increased extracellular space between fiber cells, leading to a loss of proper fiber cell shape, size, and order necessary for lens optical quality. The generation of Aqp0-deficient mice reported here has provided a valuable model system to further elucidate the vital role(s) of this MIP in lens structure and function.


    ACKNOWLEDGMENTS
 
This work was supported by National Institutes of Health Grants EY-11411 (to A. Shiels), EY-09852 (to S. Bassnett), EY-06642 (to J. Kuszak), and EY-06391 (to R. Mathias).


    FOOTNOTES
 
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).

Address for reprint requests and other correspondence: A. Shiels, Dept. of Ophthalmology and Visual Sciences, Campus Box 8096, Washington Univ. School of Medicine, 660 South Euclid Ave., St. Louis, MO 63110 (E-mail: shiels{at}vision.wustl.edu).

10.1152/physiolgenomics.00078.2001.


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Bassnett S, Becker TM, and Beebe DC. Ion concentrations, fluxes and electrical properties of the embryonic chicken lens. Exp Eye Res 54: 215–224, 1992.
  2. Bassnett S, Missey H, and Vucemilo I. Molecular architecture of the lens fiber cell basal membrane complex. J Cell Sci 112: 2155–2165, 1999.[Abstract/Free Full Text]
  3. Borgnia M, Nielsen S, Engel A, and Agre P. Cellular and molecular biology of the aquaporin water channels. Annu Rev Biochem 68: 425–458, 1999.[ISI][Medline]
  4. Chou CL, Ma T, Yang B, Knepper MA, and Verkman AS. Fourfold reduction of water permeability in inner medullary collecting duct of aquaporin-4 knockout mice. Am J Physiol Cell Physiol 274: C549–C554, 1998.[Abstract/Free Full Text]
  5. Conlon RA and Rossant J. Exogenous retinoic acid rapidly induces anterior ectopic expression of murine Hox-2 genes in vivo. Development 116: 357–368, 1992.[Abstract/Free Full Text]
  6. Costello MJ, McIntosh TJ, and Robertson JD. Distribution of gap junctions and square array junctions in the mammalian lens. Invest Ophthalmol Vis Sci 30: 975–989, 1989.[Abstract]
  7. Deen PM, Verdijk MAJ, Knoers NVAM, Wieringa B, Monnens LA, van Os CH, and van Oost BA. Requirement of human renal water channel aquaporin-2 for vasopressin-dependent concentration of urine. Science 264: 92–95, 1994.[ISI][Medline]
  8. Fernald RD and Wright SE. Maintenance of optical quality during crystalline lens growth. Nature 301: 618–620, 1983.[ISI][Medline]
  9. Fernald RD and Wright SE. Optical quality during crystalline lens growth. Nature 312: 292, 1984.[ISI]
  10. Fotiadis D, Hasler L, Muller DJ, Stahlberg H, Kistler J, and Engel A. Surface tongue-and-groove contours on lens MIP facilitate cell-to-cell adherence. J Mol Biol 300: 779–789, 2000.[ISI][Medline]
  11. Francis P, Chung JJ, Yasui M, Berry V, Moore A, Wyatt MK, Wistow G, Bhattacharya SS, and Agre P. Functional impairment of lens aquaporin in two families with dominantly inherited cataracts. Hum Mol Genet 9: 2329–2334, 2000.[Abstract/Free Full Text]
  12. Froger A, Tallur B, Thomas D, and Delamarche C. Prediction of functional residues in water channels and related proteins. Protein Sci 7: 1458–1468, 1998.[Abstract/Free Full Text]
  13. Gorin MB, Yancey SB, Cline J, Revel JP, and Horwitz J. The major intrinsic protein (MIP) of the bovine lens fiber membrane: characterization and structure based on cDNA cloning. Cell 39: 49–59, 1984.[ISI][Medline]
  14. Hamann S, Zeuthen T, La Cour M, Nagelhus EA, Ottersen OP, Agre P, and Nielsen S. Aquaporins in complex tissues: distribution of aquaporins 1–5 in human and rat eye. Am J Physiol Cell Physiol 274: C1332–C1345, 1998.[Abstract/Free Full Text]
  15. Heymann JB and Engel A. Structural clues in the sequences of the aquaporins. J Mol Biol 295: 1039–1053, 2000.
  16. Kamsteeg EJ, Wormhoudt TAM, Rijss JPL, van Os CH, and Deen PMT. An impaired routing of wild-type aquaporin-2 after tetramerization with an aquaporin-2 mutant explains dominant nephrogenic diabetes insipidus. EMBO J 18: 2394–2400, 1999.[Abstract/Free Full Text]
  17. Kuszak JR, Clark JI, Cooper KE, and Rae JL. Biology of the lens: lens transparency as a function of embryology, anatomy, and physiology. In: The Principles and Practice of Ophthalmology (2nd ed.), edited by Albert DA and Jacobiec FA. Philadelphia, PA: Saunders, 1999, p. 1355–1408.
  18. Ma T, Song Y, Gillespie A, Carlson EJ, Epstein CJ, and Verkman AS. Defective secretion of saliva in transgenic mice lacking aquaporin-5 water channels. J Biol Chem 274: 20071–20074, 1999.[Abstract/Free Full Text]
  19. Ma T, Song Y, Yang B, Gillespie A, Carlson EJ, Epstein CJ, and Verkman AS. Nephrogenic diabetes insipidus in mice lacking aquaporin-3 water channels. Proc Natl Acad Sci USA 97: 4386–4391, 2000.
  20. Ma T, Yang B, Gillespie A, Carlson EJ, Epstein CJ, and Verkman AS. Severely impaired urinary concentrating ability in transgenic mice lacking aquaporin-1 water channels. J Biol Chem 273: 4296–4299, 1998.[Abstract/Free Full Text]
  21. MacGregor G, Zambrowicz BP, and Soriano P. Tissue non-specific alkaline phosphatase is expressed in both embryonic and extraembryonic lineages during mouse embryogenesis but is not required for migration of primordial germ cells. Development 121: 1487–1496, 1995.[Abstract/Free Full Text]
  22. Mathias RT, Baldo GJ, and Rae JL. Physiological properties of the normal lens. Physiol Rev 77: 21–50, 1997.[Abstract/Free Full Text]
  23. Nemeth-Cahalan KL and Hall JE. pH and calcium regulate the water permeability of aquaporin 0. J Biol Chem 275: 6777–6782, 2000.[Abstract/Free Full Text]
  24. Piatigorsky J. Lens differentiation in vertebrates: a review of cellular and molecular features. Differentiation 19: 134–153, 1981.[ISI][Medline]
  25. Preston G, Smith BL, Zeidel ML, Moulds JJ, and Agre P. Mutations in aquaporin-1 in phenotypically normal humans without functional CHIP water channels. Science 265: 1585–1587, 1994.[ISI][Medline]
  26. Shiels A and Bassnett S. Mutations in the founder of the MIP gene family underlie cataract development in the mouse. Nat Genet 12: 212–215, 1996.[ISI][Medline]
  27. Shiels A and Griffin CS. Aberrant expression of the gene for lens major intrinsic protein in the CAT mouse. Curr Eye Res 12: 913–921, 1993.[ISI][Medline]
  28. Shiels A, Mackay D, Bassnett S, Al-Ghoul K, and Kuszak J. Disruption of lens fiber cell architecture in mice expressing a chimeric AQP0-LTR protein. FASEB J 14: 2207–2212, 2000.[Abstract/Free Full Text]
  29. Sidjanin DJ, Parker-Wilson DM, Neuhauser-Klaus A, Pretsch W, Favor J, Deen PMT, Ohtaka-Maruyama C, Lu Y, Bragin A, Skach WR, Chepelinsky AB, Grimes PA, and Stambolian DE. A 76-bp deletion in the Mip gene causes autosomal dominant cataract in Hfi mice. Genomics 74: 313–319, 2001.[ISI][Medline]
  30. Sivak JG, Stuart DD, Herbert KL, Van Oostrom JA, and Segal L. Optical properties of the cultured bovine ocular lens as an in vitro alternative to the Draize eye toxicity test: Preliminary validation for alcohols. Toxicol Meth 2: 280–294, 1992.
  31. Trokel S. The physical basis for transparency of the crystalline lens. Invest Ophthalmol Vis Sci 1: 493–501, 1962.
  32. Varadaraj K, Kushmerick C, Baldo GJ, Bassnett S, Shiels A, and Mathias RT. The role of MIP in lens fiber cell membrane transport. J Membr Biol 170: 191–203, 1999.[ISI][Medline]
  33. Yancey SB, Koh K, Chung J, and Revel JP. Expression of the gene for main intrinsic polypeptide (MIP): Separate spatial distribution of MIP and crystallin gene transcripts in rat lens development. J Cell Biol 106: 705–714, 1988.[Abstract]
  34. Yang B, Gillespie A, Carlson EJ, Epstein CJ, and Verkman AS. Neonatal mortality in an aquaporin-2 knock-in mouse model of recessive nephrogenic diabetes insipidus. J Biol Chem 276: 2775–2779, 2001.[Abstract/Free Full Text]
  35. Zambrowicz BP, Friedrich GA, Buxton EC, Lilleberg SL, Person C, and Sands AT. Disruption and sequence identification of 2,000 genes in mouse embryonic stem cells. Nature 392: 608–611, 1998.[ISI][Medline]
  36. Zampighi GA, Simon SA, and Hall JE. The specialized junctions of the lens. Int Rev Cytol 136: 185–225, 1992.[ISI][Medline]