Expression and Regulation of α-, β-, and γ-Crystallins in Mammalian Lens Epithelial Cells

Abstract

purpose. In the mammalian lens, the expression of the β- and γ-crystallin families is thought to be limited to fiber cells. However, several studies detected these proteins or their mRNAs in human lens epithelial cells. To resolve this apparent discrepancy, 14 crystallin mRNAs were examined and the expression and subcellular distribution of selected crystallin proteins in lens epithelial cells determined.

methods. Transcript levels were analyzed by quantitative real-time PCR using mRNA from P3 rat lens epithelia cultured for 0 or 20 hours or 4 or 7 days in basal medium or with added FGF2. Antibodies to βB1-, γS-, αA-, and αB-crystallins were used for Western blot analysis of proteins extracted from adult mouse, human, bovine, rabbit, and rat lens epithelial and fiber cells. Rat lenses or lens epithelia were rapidly fixed in situ, 30 minutes after death, or after dissection from the lens, and the intracellular distributions of crystallins were examined by immunostaining and confocal microscopy.

results. Four patterns of crystallin gene expression were detected in cultured lens epithelia. Transcripts encoding most β- and γ-crystallins were detectable and, in some cases, abundant at the time of explantation. Changes in crystallin protein levels in P3 epithelia cultured in basal or FGF-supplemented medium generally reflected the changes in their mRNAs. βB1- and γS-crystallins were abundant in adult human, mouse, rat, rabbit, and bovine lens epithelial cells. The α-, β- and γ-crystallins were found in distinct subcellular locations in adult lens epithelial cells. These proteins dramatically relocalized during fiber cell differentiation and after death and/or dissection of the lens epithelium.

conclusions. βB1- and γS-crystallins are normally abundant in adult mammalian lens epithelial cells. Complex programs of transcription and degradation regulate the accumulation of crystallin mRNAs in lens epithelial cells after stress, at different ages, and during cell differentiation. Because crystallins selectively localize in distinct subcellular compartments during differentiation or stress, they may function to protect lens cells from injury. After stress, most αA- and αB-crystallin subunits are not in the same macromolecular complexes.

All vertebrate lenses express high levels of crystallins belonging to two protein families. There are two α-crystallins, αA and αB, and a larger number of β- and γ-crystallins that are members of the family of small heat shock proteins. Members of the β- and γ-crystallin families share the same core tertiary structure and are often referred to as belonging to the βγ family, which includes a diverse group of non-lens proteins with a similar tertiary structure. ¹ The lenses of many species also express taxon-specific crystallins, which are enzymes that have been recruited during evolution for high-level expression in the lens. ²  

Crystallins mainly function as structural proteins, accumulating to high concentrations in the lens fiber cells, thereby generating the high refractive index and transparency needed for lens function. ³ Recently, targeted gene deletion revealed that αA-crystallin also protects lens epithelial cells against apoptosis in vivo ⁴ and in vitro ^{5 6} and that αB-crystallin appears to be important for maintaining the stability of the epithelial cell genome, at least in cultured lens cells. ⁷ Nonrefractive functions for members of the βγ family have not been identified in lens cells. 

There is a long-standing view that, in mammals, the α-crystallins are expressed in lens epithelial and fiber cells, whereas expression of the β- and γ-crystallins is restricted to fiber cells. ³ This view originated with the pioneering immunocytochemical studies by McAvoy, ⁸ who found that, in neonatal rat lenses, antibodies to α-crystallin stained both epithelial and fiber cells, but antibodies to β- or γ-crystallins stained only fiber cells. Many subsequent investigators used antibodies prepared against β- or γ-crystallin fractions to define fiber cell differentiation from cultured lens epithelial explants that had been stimulated to differentiate with growth factors. ^{9 10 11 12 13 14 15 16} Other studies showed that mRNAs encoding β- and/or γ-crystallins were not detected in neonatal lens epithelial cells, but increased greatly when these cells were stimulated to differentiate into lens fiberlike cells by treatment with fibroblast growth factors. ^{15 17 18 19}  

In spite of the prevailing view that β- and γ-crystallins are expressed only in lens fiber cells, investigators in prior studies have reported detecting β-crystallins in primary cultures of neonatal human lens epithelial cells and in cultures of immortalized neonatal human lens epithelial cells. ^{20 21} However, because crystallins were only examined after culture of these cells for at least 2 weeks in serum-containing medium, it is possible that β-crystallin expression was initiated in vitro. More recent analyses detected β- and γ-crystallin mRNAs and proteins in preparations of adult human lens epithelial cells. ^{20 22 23 24 25} Because these later studies were not designed to address crystallin localization in the intact lens, it is possible that the samples examined were contaminated with small amounts of fiber cell cytoplasm or that the antibodies used for protein detection were not specific. 

Our laboratory detected β- and γ-crystallin mRNAs in a microarray study of the transcripts present in neonatal and adult rat lens epithelia, again raising the possibility that β- and γ-crystallins are expressed in lens epithelial cells. To resolve this question, we examined the expression of 14 crystallin genes and the levels of four crystallin proteins in cultured neonatal rat lens epithelia, in adult rat lens epithelial and fiber cells, and in the epithelial cells of several species of adult mammals. The results of these studies showed that several β- and γ-crystallin mRNAs and at least two proteins from the βγ family are present and usually abundant in adult mammalian lens epithelial cells. Studies of explant cultures demonstrate that the 14 crystallin genes are regulated in one of four distinct patterns in response to culture in basal medium or medium supplemented with FGF2. Changes in the expression and subcellular distribution of the crystallins during fiber cell differentiation or after death were consistent with the possibility that these proteins protect lens epithelial cells from stress. 

Methods

Isolation and Culture of Lens Epithelia

Animal use conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and the Washington University Animals Studies Committee approved the animal protocols. Human lenses were obtained from a local eye bank. Use of human lenses conformed to the Declaration of Helsinki, and the Washington University Human Studies Committee approved the human tissue protocols. 

P3 rats were killed by decapitation, and lens epithelia were isolated by dissection in CO₂-independent medium (Invitrogen Corp., Carlsbad, CA) and spread out and pinned on 13-mm round plastic coverslips (Sarstedt, Inc., Newton, NC). The epithelial cells readily separated from the fiber cells, usually leaving an intact fiber mass. The edges of the epithelial explants were carefully trimmed with a scalpel to remove any lens fiber cells that remained attached at the margin of the explants. ²⁶ RNA was isolated immediately (T0) or after culture in 5% CO₂ and 95% air for 20 hours or for 4 or 7 days in Dulbecco’s minimal essential medium (DMEM) containing 0.1% bovine serum albumin (BSA) (basal medium; Sigma-Aldrich, St. Louis, MO) or in the same medium supplemented with FGF2 (100 ng/mL; PeproTech, Inc., Rocky Hill, NJ). 

Adult rat, rabbit, bovine, human, and mouse lenses were removed from the eye and placed epithelium side down on a dry plastic coverslip. A small drop of basal medium was applied to the lens to keep it from drying out. The posterior capsule was grasped with sharp forceps and the forceps were used to peel segments of the posterior capsule to the lens equator. After the equatorial epithelial cells had been separated from the fiber mass, the fiber mass was removed, and the epithelial sheet was flattened on the coverslip. The peripheral epithelial cells were removed with a scalpel to avoid contamination with fiber cells. The absence of contamination by fiber cells or fragments of fiber cells was confirmed by confocal microscopy and Western blot analysis for the fiber-specific major intrinsic polypeptide (MIP). Groups of four lens epithelia were lysed in radioimmunoprecipitation assay (RIPA) buffer containing a protease inhibitor cocktail (Complete; Roche Diagnostics, Indianapolis, IN) and sodium orthovanadate (0.2 mg/mL), and the lysates were stored at −80°C. 

Reverse Transcription and Quantitative Real-Time PCR

RNA was extracted from groups of eight P3 rat lens epithelial explants (RNeasy Mini Kit; Qiagen, Inc., Valencia, CA). Contaminating DNA was removed from the RNA (DNA Free kit; Ambion, Inc., Austin, TX). One microgram of total RNA was reverse transcribed at 42°C for 1 hour in a total volume of 20 μL (RETROscript kit; Ambion), with random decamers used as primers. At the end of the reaction, the mixture was heated at 92°C for 10 minutes to inactivate the reverse transcriptase. For quantitative PCR, 1 μL of cDNA was amplified for 45 cycles a master mix (iQ SYBR Green Supermix; Bio-Rad Laboratories, Hercules, CA) in a thermocycler (iCycler iQ Real-Time Detection System; Bio-Rad). The PCR conditions were 95°C for 3 minutes and 45 cycles of 95°C for 30 seconds, 58°C for 30 seconds, and 72°C for 30 seconds. Software provided with the thermocycler was used to separate the fluorescent signal due to specific products from that generated by primer dimers. Known amounts of a plasmid encoding the adenosine triphosphate (ATP)-dependent calcium channel ATP2C1 (GenBank ID: NM 131907; http://www.ncbi.nlm.nih.gov/Genbank; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD) were used to generate a standard curve. The level of fluorescence generated by each PCR product was then compared with this standard. In each assay, PCR products were analyzed by gel electrophoresis to confirm that they were of the expected size. The primers used for the PCR reactions are listed in Table 1 . Experiments were performed on tissues isolated in at least two separate experiments. In each experiment, cDNAs were assayed in two quantitative (q)PCR runs. In each qPCR run, samples were analyzed in triplicate, yielding six determinations for each cDNA sample. Expression levels were compared to the levels of ATP2C1, and the ratio was plotted (±SD). 

Western Blot Analysis

Protein concentration was measured with a protein assay reagent (D_C; Bio-Rad) in microtiter plates with BSA as a standard. Equal amounts of protein (5 μg) were loaded in each lane. Blots were stained with ponceau S to confirm equal protein loading. Proteins were separated on 12% Tris-glycine gels (Novex; Invitrogen) and transferred to nitrocellulose membranes by standard methods. To test whether the epithelial explants from adult lenses had detectable contamination by fiber cells, blots of proteins extracted from epithelial explants and fiber cells from the same lenses were probed with a rabbit polyclonal antibody to the lens fiber–specific membrane protein MIP. These samples were also tested to determine the distribution of γS-crystallin in these tissues. To determine the distribution of crystallins in lens epithelial and fiber cells, we probed blots with primary antibodies to αA- (mouse, 1:1000), αB- (rabbit, 1:2000), βB1- (rabbit, 1:250), and γS- (rabbit, 1:5000) crystallins, and bands were visualized using horseradish peroxidase-labeled anti-rabbit or anti-mouse secondary antibodies and chemiluminescence detection (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). After each reaction, the blots were stripped (Restore Western Blot Stripping Buffer; Pierce, Rockford, IL) and reprobed successively with each primary antibody to assure that the protein levels could be compared directly. Each experiment was performed a minimum of three times. 

To estimate the relative abundance of each crystallin in lens epithelial and fiber cells, serial dilutions of fiber cell total proteins and an equal amount of epithelial cell total proteins from the lenses of 8-month-old rats were prepared and analyzed by Western blot analysis. The blots were scanned, and the intensity of the bands obtained at each dilution of the fiber cell extract was compared to the intensity of the band in the epithelial cell extract. Larry David (Oregon Health Science University) kindly provided us with the percentage composition of each crystallin in adult rat cortical lens fiber cells. Using these data, we calculated the approximate percentage composition of each crystallin in adult rat lens epithelial cells. 

Immunofluorescence Analysis

Adult rats were anesthetized with ketamine and medetomidine. To examine the distribution of crystallins in epithelial and cortical fiber cells, the cornea and iris were carefully removed, and the anterior of the lens was fixed in situ for 5 minutes with 10% neutral buffered formalin. The lens was then removed from the eye, fixed for an additional hour in the same fixative, and washed in PBS. The posterior of the fixed lens was punctured with fine forceps, and the hard ball of central fiber cells was carefully dissected from the softer cortical fibers and discarded. The remaining lens cortex and epithelium were then cut into 100-μm sections parallel to the visual axis using a tissue slicer (OTS-4000; Electron Microscopy Sciences, Fort Washington, PA) and stained with antibodies to βB1 or γS-crystallin. 

To examine the distribution of crystallins in the isolated lens epithelium, adult rats were anesthetized, the cornea and iris of one eye were removed rapidly, and the anterior surface of the lens was flooded for 30 seconds with a non–cross-linking fixative (Histochoice; AMRESCO, Solon, OH), This fixative was used because it allowed the lens epithelium to be dissected from the fiber mass after fixation. We were not able to perform this dissection after even brief fixation with formalin. The lens was removed from the eye and the epithelium isolated as described. The animal was then killed by CO₂ inhalation, and 30 minutes later the contralateral lens was fixed and the lens epithelium isolated in an identical manner. In a third group, animals were killed by CO₂ inhalation, the lens was removed and the lens epithelium isolated. The epithelial explant was immediately fixed for 30 seconds with the fixative (Histochoice; AMRESCO). After isolation, explants were fixed in 10% neutral buffered formalin (Fisher Scientific, Fairlawn, NJ) for 1 hour. 

For immunostaining, lens sections or epithelial explants were washed in PBS, incubated with 10% goat serum in PBS containing 0.1% Triton X-100 for 30 minutes, and incubated overnight at 4°C with primary antibody. After three 10-minute washes in PBS, the sections were incubated with Alexa Fluor 633 goat anti-rabbit or anti-mouse IgG (Molecular Probes, Eugene, OR) at a dilution of 1:1000 for 2 hours. Sections and explants were washed with PBS, mounted in a 1:1 dilution of PBS and anti-fade medium (Vectashield; Vector Laboratories, Burlingame, CA) and viewed with a confocal microscope (Carl Zeiss Meditec, Inc., Thornwood, NY). 

Primary antibodies used included a monoclonal antibody to αA-crystallin (1:500; antibody kindly provided by Usha Andley from hybridoma cells isolated by Paul FitzGerald), a polyclonal rabbit antibody to αB-crystallin (1:200; StressGen Biotechnologies Corp., Victoria, British Columbia, Canada), a rabbit polyclonal antibody to βB1-crystallin (1:100; gift of Joseph Horwitz, Jules Stein Eye Institute, University of California Los Angeles), a rabbit polyclonal antibody to γS-crystallin (1:500; kindly provided by Graeme Wistow, Section on Molecular Structure and Function, National Eye Institute, National Institutes of Health, Bethesda, MD) ²⁷ and a rabbit polyclonal antibody to MIP (kindly provided by Alan Shiels). Nuclei were visualized after staining with TOTO-1 (100 nM; Molecular Probes, Inc.). 

Results

Expression of Crystallin Transcripts in P3 Rat Lens Epithelia

In pilot studies, we used gene microarrays (Affymetrix, Santa Clara, CA) in experiments that were intended to identify the genes that increased in expression early in the formation of myofibroblasts from neonatal rat lens epithelial cells. We were surprised to find that β- and γ-crystallin transcripts were present in the lens epithelial explants used in this study. Because these studies were not designed to determine the distribution of crystallin mRNAs in epithelial and fiber cells, we initiated separate experiments to examine whether crystallin mRNAs are normally expressed in mammalian lens epithelial cells. For these analyses, explants were carefully dissected to avoid contamination of epithelial explants with fiber cells. Real-time, quantitative RT-PCR was used to quantify the expression of 14 crystallin genes in P3 rat lens epithelia at the time of isolation or after culture in basal tissue culture medium or in medium supplemented with FGF2 (100 ng/mL). 

Experiments were performed to determine an appropriate internal standard against which to compare the levels of crystallin gene expression. The level of β-actin transcripts is often used as an internal standard for mRNA quantification. However, previous studies found that β-actin levels may vary in cells cultured under different conditions ²⁸ and our preliminary microarray results showed that the levels of β-actin mRNA varied greatly in different lens regions and in lens epithelia cultured in basal medium (Fig. 1A 1B) . We analyzed microarray data from 19 different samples of lens tissues to identify a transcript that showed minimum variation under different culture conditions and in different regions of the lens. We had microarray data for mRNA from P3 and adult rat lens epithelia, P3 rat lens epithelial explants cultured in basal medium, or medium supplemented with FGF2 or vitreous humor, whole P3 lenses, central or peripheral adult rat lens epithelial cells, and adult fiber cells. We selected the mRNA encoding an ATP-dependent calcium channel, ATP2C1, as a standard because it was one of the least variable of the more than 20,000 transcripts assayed in each of the 19 samples (Fig. 1A) . The sequence encoding this channel was cloned and used as an internal standard for all subsequent analyses. Analysis of β-actin transcripts relative to ATP2C1 confirmed that levels of β-actin mRNA increased four- to fivefold after culture of P3 rat lens epithelia for 20 hours in basal tissue culture medium, although β-actin levels were more stable after longer culture intervals (Fig. 1B) . 

Lens epithelia from P3 rat lenses were dissected, and the periphery of the tissue was trimmed to remove any contaminating fiber cells, a process that also flattened and attached the explants to the surface of the plastic coverslips on which they were dissected. ²⁶ Total RNA was extracted immediately (T0) or explants were cultured in basal medium or the same medium supplemented with 100 ng/mL of FGF2 for 20 hours, 4 days, or 7 days before RNA extraction. Extracted mRNA was reverse transcribed and crystallin transcripts were quantified by qPCR. The results of these analyses are shown in Figure 2 . 

We detected four distinct patterns of crystallin gene expression. Transcripts encoding βB1- to βB3- and γC- to γE-crystallins were detectable and, in some cases abundant in the lens epithelium at explantation (T0). Transcripts in this class decreased during culture in basal medium to nearly undetectable levels, but increased greatly after treatment for 4 or 7 days with FGF2 (Fig. 1) . Although the mRNAs encoding βA1/A3-, βA4-, γA-, and γB-crystallins were too low to detect reliably at T0, these transcripts increased markedly after 4 or 7 days of culture in medium supplemented with FGF2 and were therefore included with the other crystallin transcripts in this group. Consistent with previous reports, ^{18 26} the β-crystallin transcripts increased markedly after 4 days of culture with FGF2, whereas most of the increase in the mRNAs encoding the γ-crystallins occurred later, between 4 and 7 days of culture. For the mRNAs that were detectable at T0, transcript levels decreased after 20 hours of culture, whether in DMEM or FGF-supplemented medium. This decrease was always greater in explants treated with FGF2 than in explants cultured in basal medium. Although the decrease in some of the β- and γ-crystallin mRNAs may be accounted for by the loss in culture of a low level of contaminating fiber cells, we believe that the care that we took to exclude fiber cell contamination, the increases in other mRNAs, and the acceleration of the rate of mRNA loss by FGF, a known survival factor, makes this interpretation less likely. The second pattern of expression was seen for βA2- and γS-crystallin mRNAs. These transcripts were low at T0, but increased substantially in basal medium. Twenty hours after explantation, lower levels of these mRNAs were present in explants treated with FGF2 than in explants cultured in basal medium. Treatment with FGF2 for 4 or 7 days caused the accumulation of high levels of these messages. Transcripts encoding αA- and αB-crystallins represented the third and fourth patterns of crystallin gene expression, respectively. Levels of both transcripts were high at T0, but αA-crystallin mRNA levels declined by about two thirds in basal medium and increased after treatment with FGF2, whereas αB-crystallin transcripts increased substantially in basal medium, but declined after 7 days of culture in medium supplemented with FGF2. 

Accumulation of Crystallin Proteins in Mammalian Lens Epithelial Cells

We used antibodies against selected α-, β- and γ-crystallins to determine the levels of crystallin protein accumulation in lens epithelial cells. We received an anti-peptide antibody (Lap 20) produced by Joseph Horwitz that had been prepared against a synthetic peptide at the carboxyl terminus of bovine βB1-crystallin. This antibody has been tested and shown to be specific for human and rabbit βB1-crystallin by Western blot analysis and confocal microscopy, to the extent that it was used to show that βB1-crystallin is normally expressed in bovine and rabbit ciliary epithelial cells. ²⁹ Rat, mouse, and human βB1-crystallins have identical amino acid sequences in the region recognized by this antibody. We also obtained an anti-peptide antibody prepared to a region of γS-crystallin that is conserved in γS-crystallins and not found in other crystallins and generated previously in the laboratory of Graeme Wistow. ²⁷ The specificity of this antibody was confirmed by probing Western blot analysis of total proteins from the lenses of wild-type and γS-crystallin–knockout mice (kindly provided by Graeme Wistow) (Fig. 3A) . To assure that relative levels of the different crystallins were directly comparable in each experiment, Western blots were probed with one antibody, then successively stripped and reprobed with each of the other antibodies. 

Lens fiber cells represent a potential source of crystallin contamination during mRNA isolation or Western blot analysis. Therefore, epithelial explants were carefully dissected and thoroughly examined after dissection for the presence of fiber cells. The edges of the explants, which can contain fiber cells at early stages of differentiation, were carefully trimmed. Lens epithelial explants were also stained with Alexa488-labeled phalloidin, a fluorescent probe that strongly stains the actin filaments in fiber cells. This approach revealed no fiber cell contamination of the epithelial explants (data not shown). The extensive examination with confocal microscopy of dozens of explants stained with antibodies to αA-, αB-, βB1-, and γS-crystallins, as shown in Figure 8 , also revealed no contamination with fiber cells, fragments of fiber cells, or soluble crystallins. To further test for fiber cell contamination of epithelial preparations, lens fiber and epithelial proteins from adult rats were analyzed by Western blot analysis with an antibody against the fiber-specific membrane protein MIP (aquaporin 0). A strong band was present at the appropriate molecular weight in fiber cells, but not in epithelial cells, even after overexposure of the blot (Fig. 3B) . In contrast, strong bands that reacted with the antibody to γS-crystallin were present in epithelial and fiber cells in these same extracts (Fig. 3B) . These tests demonstrate that our lens epithelial cell preparations were not contaminated with fiber cell proteins and that γS-crystallin is expressed in adult rat lens epithelial cells. 

Lens epithelial explants from P3 rat lenses were analyzed by Western blot analysis immediately after dissection or after culture for 7 days in basal medium or medium supplemented with FGF2 (Fig. 4) . P3 lens epithelial cells expressed detectable levels of αA-, αB- and βB1-crystallins, but γS-crystallin was not detectable at this age. Culture for 7 days in medium supplemented with FGF2 led to the accumulation of large amounts of all four crystallins (Fig. 4) . However, these proteins responded differently to culture in basal medium. Levels of αA- and βB1-crystallins were unchanged or increased slightly, whereas αB-crystallin was of equal or greater abundance in explants cultured in basal medium as in medium containing FGF. In agreement with changes in the expression of γS-crystallin mRNA, γS-crystallin protein, which was undetectable in P3 lens epithelial cells, accumulated to detectable levels in basal culture medium. 

Western blot analysis of lens epithelial proteins isolated at several postnatal ages showed that, with increasing age, the two α-crystallins accounted for an increasing fraction of epithelial cell proteins (Fig. 5A) . Levels of βB1-crystallin increased steadily between P3 and P28 and then decreased slightly in adult lens epithelia. γS-crystallin levels increased at each postnatal age analyzed. 

Western blot analysis of proteins extracted from the fiber cells of the same lenses as the epithelial cells analyzed in Figure 5A showed that αA-, αB-, and βB1-crystallins accounted for a similar fraction of total fiber cell protein at all ages (Fig. 5B) . However, as reported previously, ^{3 17 18 32} γS-crystallin represented an increasing proportion of total fiber cell proteins throughout postnatal development. To compare the levels of these crystallins in lens epithelial and fiber cells from adult rats, we analyzed the same amount of total protein from epithelial and peripheral fiber cells by Western blot analysis (Fig. 6A) . 

A more accurate comparison of the concentration of α-, β-, and γ-crystallins in lens epithelial and fiber cells was made by performing Western blot analysis on serial dilutions of fiber cell proteins and comparing the staining intensity of the crystallin bands in these diluted extracts with the staining intensity of undiluted extracts of epithelial cells (Fig. 6B) . These blots were scanned to determine the concentration of each crystallin in the epithelial cells, compared with its concentration in the cortical fiber cells. Using the percent composition of the crystallins in the cortical fiber cells of adult rats, ³³ we then calculated the approximate percent composition of each protein in adult rat lens epithelial cells (Table 2) . 

To determine whether these crystallins were common components of mammalian lens epithelial cells, we performed Western blot analysis of adult lens epithelia from several species (Fig. 7) . αB-, βB1- and γS-crystallins were abundant in adult human, mouse, rat, rabbit, and bovine lens epithelial cells. High levels of αA-crystallin were present in the epithelial cells of all species except human, where levels were detectable, but lower than in the other mammals. We do not know whether the relatively low level of αA-crystallin in human lens epithelial cells is a consequence of the older age of the human donors that we tested or whether it reflects a more fundamental difference in the regulation of αA-crystallin in human lens epithelial cells compared with the other mammals in our survey. Other investigators found a marked predominance of αB- over αA-crystallin in human lens epithelial proteins that had been separated by two-dimensional gel electrophoresis and stained with Coomassie blue (Joseph Horwitz, personal communication, October 2003). We cannot explain the slower migration of βB1-crystallin from bovine lens epithelia, since the cDNA sequence for βB1-crystallin suggests that this protein should be of similar size as βB1-crystallin in other mammalian species. Blots using antibodies to βB1-crystallin frequently showed a more rapidly migrating band (see also Figs. 5 6 ), consistent with previous reports of N-terminal truncation of this protein. ^{34 35}  

The Subcellular Distribution of α-, β-, and γ-Crystallins in Adult Lens Epithelial Cells

The distributions of selected crystallins in adult rat lens epithelial and fiber cells were determined using immunofluorescence and confocal microscopy on sections of whole lenses and in wholemounts of isolated epithelial explants (Figs. 8 9) . Because Western blot analysis for βB1- and γS-crystallins indicated these proteins should be present in adult lens epithelial cells, we tested their distribution in sections of whole lenses (Fig. 8) . Rats were anesthetized, the cornea and iris were removed, and the lens was fixed in situ. Lenses were then removed from the eye, sectioned and stained with antibodies to βB1- and γS-crystallins. βB1-crystallin was present in the cytoplasm of cells throughout the lens epithelium, with low levels of staining detectable in the nuclei. A striking redistribution of this protein occurred as the epithelial cells began to differentiate into fiber cells at the lens equator. Nuclear staining for βB1-crystallin increased greatly, so that the nuclei of the fiber cells stained more strongly than the cytoplasm. 

Staining for γS-crystallin in epithelial and fiber cells followed a reciprocal pattern to that observed for βB1. Epithelial cell nuclei were strongly stained by the antibody to γS-crystallin, whereas levels of this protein were lower in the cytoplasm. At the lens equator, γS-crystallin staining decreased markedly in the nuclei, and nearly all the staining became cytoplasmic. These observations show that the subcellular distribution of these crystallins is differentially regulated during lens cell differentiation. 

We used wholemounts of lens epithelia to explore further the subcellular localization of the α-, β-, and γ-crystallins. In preliminary studies, we detected variable distributions of some of these crystallins between the cytoplasm and the nucleus. At least part of this variability was traced to differences in the interval between death and fixation and to the stress associated with the isolation of epithelial explants. To address these sources of variation, we anesthetized adult rats, removed the cornea and iris of one eye rapidly, and fixed the lens epithelial cells while the lens was still in the eye. The animal was then killed by CO₂ inhalation, and 30 minutes later the contralateral cornea and iris were removed, and the lens epithelial cells were fixed in situ. After fixation, the lens epithelia were dissected from the lens and prepared for immunostaining. In another group of animals, death was induced by CO₂ inhalation, the lens was removed from the eye, the epithelium was dissected, explanted onto a plastic coverslip and then fixed. These protocols provided sheets of lens epithelial cells that were fixed immediately in situ, 30 minutes after death, or after death and dissection. 

Significantly different intracellular distributions of the four crystallins were seen after these treatments (Fig. 9) . In epithelial cells fixed in situ in the living animal, αA-crystallin immunoreactivity localized predominantly at the plasma membrane and in a fibrillar distribution throughout the cytoplasm, with little nuclear staining. Thirty minutes after death, αA-crystallin was distributed in a pattern that was similar to cells fixed before death, although the cytoplasmic staining was less fibrillar and more granular. When lens epithelia were fixed after dissection, most of the αA-crystallin immunoreactivity was at the plasma membrane. The remainder of the staining was in coarse cytoplasmic granules. In epithelial cells that were fixed in the living eye, the distribution of αB-crystallin immunostaining was similar to that of αA-crystallin, with fibrillar cytoplasmic staining, moderate staining at the plasma membrane, and little nuclear staining. Thirty minutes after death, αB-crystallin immunoreactivity increased at the plasma membrane and became granular in the cytoplasm. There was a marked redistribution of αB-crystallin in epithelial cells fixed after dissection, with staining diffusely distributed throughout the cytoplasm and in large, intensely stained globules. In epithelial cells fixed before death, βB1-crystallin was predominantly found in the cytoplasm in fine granules against a background of diffuse, uniform staining. A substantial amount of staining for βB1-crystallin was also evenly distributed within the nuclei. After death, βB1-crystallin showed fibrillar and fine granular distribution in the cytoplasm. Most nuclei had little staining. When epithelial cells were fixed after dissection, βB1-crystallin immunoreactivity appeared in fine cytoplasmic granules, in a distinct perinuclear distribution, and in the nucleus as fine granules against a diffuse background. In lens epithelial cells fixed in situ, most of the γS-crystallin immunostaining was present as fine granules in the nucleus. A smaller number of granules were distributed uniformly throughout the cytoplasm. After death, most of the γS-crystallin localized to the nucleus, with only a few coarse granules remaining in the cytoplasm. When explants were dissected before fixation, γS-crystallin staining was found entirely in the nucleus. 

Discussion

The accumulation of α-, β-, and γ-crystallins in lens fiber cells is important for the refractive properties of the lens. ³ The α-crystallins protect other lens proteins from denaturation and aggregation ^{3 36 37} and can protect lens epithelial cells from stress-induced cell death. ^{4 5 6} The data presented demonstrate that at least two members of the βγ-crystallin family are also abundant in the epithelial cells of adult mammalian lenses, raising the possibility that these proteins have nonrefractive functions in lens cells. 

Because fiber cells express very high levels of crystallins, we were careful to rule out the contamination of lens epithelial preparations by mRNAs or proteins from fiber cells. These steps included careful examination of epithelial explants during dissection for evidence of contamination by fiber cells and trimming the edges of epithelial explants to remove fiber cells at early stages in their differentiation. We did not detect fiber cells or fiber cell fragments when lens epithelial preparations were examined by confocal microscopy after being stained with antibodies to specific α-, β-, or γ-crystallins or with fluorescent derivatives of the filamentous actin-binding protein, phalloidin. Tests for fiber cell contamination in extracts of epithelial cells failed to reveal detectable amounts of the fiber-specific membrane protein MIP. These precautionary measures and the straightforward detection of βB1- and γS-crystallins in adult lens epithelial cells by Western blot analysis and confocal microscopy demonstrate that these crystallins are normal components of adult mammalian lens epithelial cells. 

Given the intense interest in crystallin gene expression over the past 30 years, it is appropriate to ask why β- and γ-crystallin mRNAs and proteins were thought not to be present in mammalian lens epithelial cells. There are probably several explanations: (1) Although we found that γS- and βB1-crystallin were abundant in adult mammalian lens epithelial cells, levels of γS- and βB1-crystallin were undetectable and low, respectively, in neonatal rat lens epithelia. Because most in vitro assays of fiber cell differentiation were performed on neonatal rat lens epithelia, ^{9 10 14 26} these studies would have been unlikely to detect these proteins. Because we did not test antibodies to other individual β- or γ-crystallins, we cannot be sure whether other crystallin proteins are present in neonatal rat lens epithelial cells. However, we did detect substantial levels of transcripts for all members of the βB-crystallin group and for γC-crystallin in P3 rat lens epithelial cells (Fig. 2) . This observation and our data showing that βB1-crystallin is present at low concentrations in epithelial cells at this age suggest that these proteins are present in the neonatal rat lens epithelial cells, but were not detected previously. (2) In previous studies, β- and γ-crystallins were typically detected using antibodies raised to mixtures of these proteins. The sensitivity of these antibodies to a single crystallin is unknown. If the antibodies reacted preferentially with a crystallin that is present at substantially higher levels in fiber cells than in epithelial cells, it may have appeared that an entire class of crystallin was “absent” from lens epithelial cells. The antibodies to specific crystallins that were used in the present study provided more reliable quantification and localization. In addition, immunocytochemistry is not the preferred method for deciding whether a protein is absent from a tissue, since it is difficult to discriminate between nonspecific background staining and specific staining of a low-abundance antigen. (3) Some of the mRNAs detected in neonatal lens epithelia in this study were quantified by qPCR, a very sensitive method of analysis. Previous studies using in situ hybridization ¹⁷ or Northern blot analysis ^{15 18} may not have been sufficiently sensitive to detect these molecules at the low levels that may be present in freshly isolated neonatal lens epithelial cells. (4) Treatment of P3 rat lens epithelia with FGF2 caused a large increase in the levels of nearly all crystallin mRNAs and proteins. If one were to compare stimulated and unstimulated samples on the same Northern blot, for example, the high signal in the stimulated sample might be overexposed before the signal in the unstimulated sample was detectable. (5) Given the widely held view that β- and γ-crystallins are not expressed in the lens epithelium and lacking the specific antibodies used in the present study, an investigator detecting low levels of these proteins in a sample might have attributed this result to contamination from lens fiber cells or to a nonspecific antibody reaction. 

Several β-crystallin mRNAs are expressed in the lens epithelia of birds. ³⁸ In chickens, β-crystallin transcripts accumulate during later stages of embryogenesis and during posthatching development. Based on the expression of βB1-crystallin in the present study, postnatal increases in the levels of β-crystallins in lens epithelial cells may be the typical pattern of crystallin gene expression in birds and mammals, rather than a taxon-specific peculiarity. 

Regulation of Crystallin Gene Expression in Cultured Lens Epithelial Explants

Our data are in accord with many studies showing that the fiberlike cells that form after treatment of neonatal rat lens epithelia with FGFs accumulate large amounts of α-, β-, and γ-crystallin mRNAs and proteins. ^{11 14 17 18 26} However, our studies detected unexpected increases in αB-, βA2-, and γS-crystallin mRNA accumulation when epithelial explants were cultured in basal tissue culture medium. In the case of αB-crystallin, mRNA levels continued to increase in basal medium up to 7 days in culture. At this time αB-crystallin mRNA was about three times more abundant in basal medium than in FGF-supplemented medium. Increases in these mRNAs were reflected in the levels of the proteins they encoded. γS-crystallin was undetectable at the time of explantation, but was detected after 7 days of culture in basal medium. After culture for 7 days in basal medium, αB-crystallin levels were equal to or greater than in FGF-supplemented medium. Because the levels of αB-crystallin mRNA increased in basal medium and decreased in the same medium supplemented with FGF2, FGF signaling must actively suppress the accumulation of αB-crystallin mRNA. These observations reveal a new mode of crystallin gene expression in cultured lens epithelial cells that is independent of FGF signaling and that appears to depend on the stress associated with explantation and culture in vitro. 

Our data show that γS-crystallin is regulated in a different manner than other γ-crystallins in rat lens epithelial cells and suggest that γS-crystallin may have functions that are different from its closest evolutionary relatives. In mammals, the genes encoding γA- to γF-crystallin are clustered in a tandem array, whereas the γS-crystallin gene is on a different chromosome. Sequence comparisons reveal greater similarity between the members of the γ-crystallin cluster than to γS-crystallin, suggesting that the genes encoding mammalian γA- to γF-crystallins arose more recently by gene duplication. ^{39 40} The γA- to γF-crystallins are expressed at high levels during embryogenesis, but expression declines rapidly after birth, ^{17 19} whereas γS-crystallin accumulation begins soon after birth and continues at high levels throughout life. ^{3 34} γS-crystallin is the most evolutionarily conserved of the γ-crystallins and all the γ-crystallin genes except γS were lost during the evolution of birds and reptiles. These distinctions suggest that γS-crystallin has a different and more fundamental function than the members of the γ-crystallin cluster. 

All the transcripts that were measurable at 20 hours after explantation were lower in cells treated with FGF than in cells cultured in basal medium. To our knowledge, such short-term changes in transcript levels have not been reported previously. Given that this effect was seen for transcripts that were accumulating (e.g., αB-crystallin) and those that were decreasing in abundance (e.g., γC-crystallin), it seems most likely that treatment with FGF causes an increase in the rate of mRNA turnover. An increase in the rate of mRNA degradation would slow the rate of accumulation and hasten the rate of decrease of transcript levels. The possible short-term effects of FGFs on mRNA turnover in lens epithelial cells may warrant further investigation, because a similar destabilization of mRNAs may occur during the early stages of fiber cell differentiation. Rapid mRNA turnover at this stage might contribute to the removal of mRNAs encoding proteins that are needed for the function of epithelial cells, but that would interfere with the function of fiber cells (e.g., the epithelium-specific connexin, Cx43 ⁴¹ or epithelium-specific ion channels ⁴² ). 

A previous study showed that αB-crystallin may exist independent of αA-crystallin in lens and nonlens tissues. ⁴³ Therefore, it would be interesting to know whether αB-crystallin normally forms hetero-oligomers with αA-crystallin, or whether these proteins function independently in epithelial cells. Our data indicate that αA- and αB-crystallins distribute to different regions of adult lens epithelial cells after dissection of the lens epithelium, indicating that, under these conditions, these proteins are not in the same oligomers. 

Subcellular Localization of Crystallins in Lens Cells

We detected major changes in the subcellular localization of crystallins during fiber cell differentiation and after death. Although the mechanisms responsible for and the significance of the changes in intracellular protein localization must await further analysis, several conclusions can be drawn from our observations. 

The subcellular redistribution of βB1- and γS-crystallin during fiber cell differentiation raises the possibility that these proteins play distinct roles in epithelial and fiber cells. This observation also suggests that there are specific mechanisms that regulate the transport of these proteins into and out of the nucleus, even though neither protein has a nuclear localization signal that can be identified by sequence analysis. 

The existence of mechanisms to regulate the subcellular distribution of crystallins is also supported by the observation that each of the α-, β- and γ-crystallins studied were differentially distributed in lens epithelial cells that were fixed in situ. These proteins then changed their subcellular distributions after death or after the dissection of epithelial explants. Therefore, there appear to be distinct mechanisms that govern the distribution of different crystallins within adult rat lens epithelial cells after stress. The dissection of the lens epithelium is likely to subject the epithelial cells to physical and biochemical stresses. Signals generated by these stresses may cause the selective redistribution of the crystallins. The application of defined physical or biochemical stimuli to lens epithelial cells in situ, followed by rapid fixation, could uncover the signaling pathways that underlie the redistribution of crystallins in stressed lens epithelial cells. 

α-Crystallins associate with the plasma membranes of normal lens fiber cells and in increasing amounts during cataract formation. ^{44 45 46 47 48 49} In the present studies, a substantial amount of αA- and αB-crystallin associated with the plasma membranes of lens epithelial cells fixed immediately or 30 minutes after death, indicating that α-crystallin also has a propensity for associating with the membranes of lens epithelial cells. In lens cells fixed after dissection, most of the αA-crystallin was associated with the plasma membrane, whereas the majority of the αB-crystallin was distributed in the cytoplasm. These results suggest that αA- and αB-crystallins have different affinities for plasma membranes in stressed lens epithelial cells. They also demonstrate that, under these conditions, most of the epithelial αA- and αB-crystallins are not in the same aggregates. 

Stress Crystallins

The α-crystallins are members of the family of small heat shock proteins. ^{1 3 50} They function as molecular chaperones that protect other proteins against stress-induced aggregation ^{3 4 36 37} and protect the epithelial cells in which they are expressed from apoptosis. ^{4 5 6} As with most other members of the family of small heat shock proteins, the level of αB-crystallin is increased by diverse stresses in several cell types. ^{5 51 52 53 54} However, heat shock or other stresses do not appear to alter the synthesis or accumulation of αA-crystallin. ^{3 52 55 56}  

The present study shows that αB-crystallin increases in lens epithelia cultured in basal medium, indicating that culture in this manner constitutes a significant stress for these cells. If this interpretation is correct, the concomitant increase in the expression of βA2- and γS-crystallin mRNAs suggests that these proteins also protect lens epithelial cells against stress. Mutations in the γS-crystallin gene in two mouse strains leads to cataracts and to defects in the lens epithelial cells, ^{57 58} further supporting a protective role for this protein in epithelial cells. The observation that γS-crystallin is predominantly found in the nucleus of lens epithelial cells and redistributes entirely to the nucleus after dissection is consistent with a role for it in protecting against DNA damage. 

 Supported by an unrestricted grant from Research to Prevent Blindness, National Eye Institute Grant EY04853 (DCB), Supplemental Grant EY05856 from the National Eye Institute to support microarray studies, and Core Grant EY02687 for Vision Research to the Department of Ophthalmology and Visual Sciences.

 Submitted for publication April 15, 2004; revised June 15, 2004; accepted July 2, 2004.

 Disclosure: X. Wang, None; C.M. Garcia, None; Y.-B. Shui, None; D.C. Beebe, None

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

 

Table 1.

 

PCR Primers Used to Quantify Crystallin mRNA Levels

Transcript	Primer Sequences
αA	5′-AGCCGACTGTTCGACCAGTTC-3′
		5′-AACTTGTCCCGGTCAGATCG-3′
αB	5′-CTTCGGAGAGCACCTGTTGG-3′
		5′-GGAGAGAAGTGCTTCACGTCCAG-3′
βA1/A3	5′-GAGAATACCCTCGATGGGATGC-3′
		5′-CCACTGGCGTCCAATAAAGTTC-3′
βA2	5′-CCAGGGACAGCAGTTCATTCTAG-3′
		5′-ATGGAAGGCAGTGATGGGTAG-3′
βA4	5′-GGCTGACCATCTTCGAGCAG-3′
		5′-GCACACTCTGCACCTGGAAGG-3′
βB1	5′-GCCAACTTCAAGGGCAACAC-3′
		5′-GCATCTGCGGCTGGAAGG-3′
βB2	5′-CTGGGTGGGCTACGAGCAG-3′
		5′-TGGCTGTCCACTTTGATGGG-3′
βB3	5′-TTGAGAACCCAGCCTTCAGTG-3′
		5′-CCGCTCGAACACGTACTGG-3′
γA	5′-ACTCCATTCGCTCCTGCCGTTC-3′
		5′-GATGCAGGAACAGTCTTCCG-3′
γB	5′-TGCTGGATGCTCTATGAGCGA-3′
		5′-GGTGGAAGCGATCCTGAAGAG-3′
γC	5′-TGCTGGATGCTCTATGAGCGA-3′
		5′-GGATGCAGGAGCAGTCTTCAC-3′
γD	5′-CTATGAGCAGCCCAACTTCACAG-3′
		5′-TGGAATCGGTCCTGGAGAGAG-3′
γE	5′-ATGAGCAGCCCAACTTCACA-3′
		5′-GAGTGGAAGTCACTGAAGTGGAAG-3′
γS	5′-CGAGTACCCTGAGTACCAGCGT-3′
		5′-CAATGCGGCGGAATGACTG-3′
ATP2C1	5′-GCACTTACCCAGCAGCAGAGAG-3′
		5′-TACAATCCCAGACGACTAGCGA-3′

Figure 1.

 

Microarray and qPCR data comparing the expression of β-actin and the ATP-dependent calcium channel ATP2C1 in lens cells. (A) The normalized signals for β-actin and ATP2C1 transcripts in 19 different microarray experiments performed using RNA extracted from different lens cell preparations. The samples analyzed by microarray were lane 1: adult rat lens epithelium, eyelid closure for 3 days; lane 2: normal adult rat lens epithelium; lane 3: adult rat lens epithelium, animal exposed to 60% oxygen for 3 days; lane 4: adult rat central lens epithelium; lane 5: adult rat peripheral lens epithelium; lanes 6 and 7: P3 rat lens epithelium; lanes 8 and 9: adult rat lens epithelium; lanes 10 and 11: P3 rat lens epithelium, cultured for 20 hours in basal medium; lanes 12 and 13: P3 rat lens epithelium cultured for 20 hours in FGF2 (50 ng/mL); lanes 14 and 15: P3 rat lens epithelium cultured in for 20 hours in 20% bovine vitreous humor; lanes 16 and 17: adult rat lens epithelium cultured for 20 hours in basal medium; and lanes 18 and 19: whole P3 rat lenses cultured for 20 hours in basal medium. (B) Determination by qPCR of the levels of β-actin mRNA, relative to ATP2C1 in P3 rat lens epithelial explants cultured for 0 hours (T0), 20 hours in basal medium (20h), 20 hours in FGF2 (20FGF), 4 days in basal medium (4D), 4 days in FGF2 (4DFGF), 7 days in basal medium (7D), or 7 days in FGF2 (7DFGF).

Figure 2.

 

Expression of 14 crystallin genes in P3 rat lens epithelial cells cultured as described in Figure 1 . The y-axis scale is different for different crystallin transcripts. For several of the crystallin transcripts, the high level of expression after culture for 7 days in FGF obscured the expression of these genes at T0. Only the mRNAs encoding βA1/A3-, βA4-, γA-, and γB-crystallins were too low in expression to detect reliably at T0. The data presented are for one representative experiment in which mRNA was isolated from eight lens epithelia at each data point. Each data point represents six qPCR determinations. Expression levels are relative to ATP2C1, an ATP-dependent calcium channel. Ratios are ±SD.

Figure 3.

 

(A) Confirmation of the specificity of the antibody to γS-crystallin used in these studies. Top: staining with the antibody to γS-crystallin in wild-type (+/+) and γS-crystallin knockout (−/−) lenses. Bottom: staining obtained in the same lens extracts with an antibody to macrophage migration inhibitory factor (MIF), a cytokine that is expressed at high levels in the lens. ^{30 31} (B) Western blot of proteins extracted from adult rat lens epithelial explants and fiber cells. Top: the fiber-specific membrane protein, MIP (aquaporin 0), was readily detected in extracts of fiber cells, but was not detectable in extracts from lens epithelia prepared in the same manner as the epithelia used in the other studies described in this article. This blot was exposed for 5 seconds. Exposing the blot for up to 2 minutes did not reveal a band for MIP in the lens epithelial sample. Bottom: in the same samples, γS-crystallin was abundant in lens epithelial and fiber cells.

Figure 4.

 

Expression of four crystallins in freshly isolated P3 rat lens epithelia (T0) and epithelia cultured for 7 days in medium supplemented with FGF2 (7F) or in basal medium for 7 days (7C). The top band in the blot stained with antibodies to αA-crystallin (cryaa) results from the translation of an alternatively spliced mRNA containing an additional exon (αA_ins). The same blot was stripped and reprobed with antibodies to each of the crystallins. Five micrograms of protein was loaded in each lane.

Figure 5.

 

Crystallin levels in rat lens epithelia (A) and fibers (B) from rats of different postnatal ages (P3–P28 and adult). Five micrograms of protein extract was loaded in each lane.

Figure 6.

 

(A) Comparison of crystallin levels in adult rat lens epithelial (LE) and fiber (LF) cells. Five micrograms of protein extract was loaded in each lane. (B) Serial dilutions to estimate the relative accumulation of crystallins in lens epithelial and fiber cells. Extracts of cortical lens fiber cells from 8-month-old rats were prepared as twofold serial dilutions (F1, F2, F4, F8, and F16). Staining for γS-crystallin was then compared between the serial dilutions and undiluted extract of lens epithelial cells (E1) from the same lenses used to prepare the fiber cell extract. The undiluted fiber and epithelial extracts had the same amount of total protein. The blots were then scanned to quantify the relative concentration of crystallins in the epithelial cells. Similar comparisons were made for αA-, αB-, and βB1-crystallins.

Table 2.

 

Crystallin Composition of Cortical Fiber and Lens Epithelial Cells

Crystallin Protein	Percent Composition in Adult Rat Cortical Fiber Cells*	Relative Concentration from Western Blot	Calculated Percent Composition in Adult Rat Lens Epithelial Cells
αA	12.6	1/6	2.1
αB	1.7	1/2	0.9
βB1	3.0	1/4	0.8
γS	7.5	1/4	1.9

*  Crystallin composition in adult rat cortical lens fiber cells was provided by Dr. Larry David from data reported in Reference ³³ . These data were obtained from 4-month-old rats, which may differ from the 8-month-old rats used in the present study. The percentage composition for γS-crystallin may be elevated by partial contamination with βA1/A3-crystallin.

Figure 7.

 

Crystallin proteins extracted from the isolated lens epithelia of five adult mammals. Five micrograms of protein extract was loaded in each lane. Ra, rat; M, mouse; H, human; Rb, rabbit; B, bovine.

Figure 8.

 

Staining of sections of whole lenses with antibodies to (top) βB1- or (bottom) γS-crystallin. βB1-crystallin immunostaining was present throughout the lens epithelial and fiber cells. In the lens epithelium, most of the staining was cytoplasmic. At the equatorial region, staining for βB1-crystallin became stronger in the nuclei than in the cytoplasm (arrow). This difference was particularly evident in the deeper fiber cell nuclei. Staining for γS-crystallin mainly localized to the nuclei of epithelial cells, although the cytoplasm was also stained. Nuclear staining for γS-crystallin decreased as the epithelial cells began to differentiate into fiber cells at the lens equator (arrow). Nuclei counterstained with TOTO-1. Scale bar, 20 μm.

Figure 9.

 

Antibody staining of lens epithelial explants that were fixed immediately, 30 minutes after death, or after the lens epithelium had been dissected from the lens. The lowest row of images shows the appearance of the nuclei of explants after staining with the nucleic acid dye TOTO-1. All images were obtained at the same magnification. Scale bar, 20 μm.

The authors especially thank Graeme Wistow, who offered many helpful suggestions; as well as Larry David and Kirsten Lampe, who provided data on the percentage distribution of crystallins in adult rat cortical fiber cells; and Usha Andley, Mark Petrash, and Steven Bassnett, who provided thoughtful comments on the manuscript.