αB-crystallin has been demonstrated, in tissue culture experiments,to be a caspase 3 inhibitor; however, no animal model studies have yet been described. Here, we show that morphological abnormalities in lens secondary fiber cells of αA-/αB-crystallin gene double knockout (DKO) mice are consistent with, and probably result from, elevated DEVDase and VEIDase activities, corresponding to caspase 3 and caspase 6, respectively. Immunofluorescence microscopy revealed an increased amount of caspase 6, and the active form of caspase 3, in specific regions of the DKO lens, coincident with the site of cell disintegration. TUNEL labeling illustrated a higher level of DNA fragmentation in the secondary fiber lens cells of DKO mice,compared with wild-type mice. Using a pull-down assay, we show interaction between caspase 6 and αA- but not αB-crystallin. These studies suggest that α-crystallin plays a role in suppressing caspase activity,resulting in retention of lens fiber cell integrity following degradation of mitochondria and other organelles, which occurs during the apoptosis-like pathway of lens cell terminal differentiation.

The vertebrate lens offers a unique model system for studying mechanisms of cell death. The mature ocular lens consists of three cell types: epithelial cells, primary lens fiber cells and secondary lens fiber cells. Primary fiber cells are generated in early embryogenesis from elongated epithelial cells at the posterior region of the lens vesicle. Removal of nuclei and other organelles from these cells occurs through accumulation of osmiophillic bodies that are later digested in vesicles with proteolytic enzymes, followed by removal of the breakdown products from the cell(Vrensen et al., 1991). Following this process, these cells remain unchanged in the central part of the lens, known as the lens nucleus. However, the lens constantly grows throughout the life of the organism, by formation of new secondary lens fiber cells from the epithelial cells at the lens equator, which differentiate and elongate. These newly formed fiber cells form an onion-like array of concentric layers. Maturation of the secondary lens fiber cells parallels many classical apoptosis events: chromatin condensation, DNA fragmentation,nucleopore clustering with subsequent nuclear breakdown, and disappearance of mitochondria and other organelles. However, lens fiber cell differentiation shows signs of attenuated or arrested apoptosis: breakdown of nuclei occurs in days, not hours; the cytoskeleton remains intact throughout the remaining life of the organism; and the lens fiber cells show increased resistance to apoptotic agents such as staurosporine and etoposide(Dahm, 1999; Wride, 2000).

Differentiation of the secondary fiber cells is accompanied by lens-specific expression of several families of crystallin genes(Piatigorsky, 1981). Crystallins are the predominant soluble proteins (20-60% of the lens wet weight) in the mammalian lens, reaching a physiological concentration of∼300 mg/ml (Harding,1991). One crystallin family found in all vertebrate species examined (Wistow and Piatigorsky,1988), is α-crystallin, which consists of two subunits,αA (Cryaa, Crya1, Hspb4) and αB (Cryab, Crya2, Hspb5) in an approximate 3:1 ratio, with 55% amino acid sequence identity. Theα-crystallins are also members of the small heat-shock protein family,and αB-crystallin is a bona fide heat shock protein.

Earlier studies described a potential role for α-crystallin as an anti-apoptotic agent because of its ability to protect the cell from hypertonic (Dasgupta et al.,1992), heat (Aoyama et al.,1993), oxidative and TNFα-induced(Mehlen et al., 1995) stress. This list of stress factors is continually expanding. α-Crystallin has been shown to prevent apoptosis induced by a variety of agents including UV radiation (Harding, 1991),TNFα (Andley et al.,2000; Mehlen et al.,1996a), staurosporine (Andley et al., 2000; Mehlen et al.,1996a; Mehlen et al.,1996b; Mao et al.,2004), hydrogen peroxide (Mao et al., 2001), sorbitol (Mao et al., 2004), etoposide (Mao et al., 2004) and okadaic acid(Li et al., 2001). Recent reports have begun to elucidate the molecular mechanisms by whichα-crystallin inhibits apoptosis. Experiments with cell extracts and in tissue culture suggest that αB-crystallin increases resistance to the apoptotic inducers etoposide and TNFα by decreasing caspase 3 activity via inhibition of autocatalytic maturation of procaspase 3(Kamradt et al., 2001; Kamradt et al., 2002). Coimmunoprecipitation of αB-crystallin with caspase 3 and its uncleaved precursor form suggests a direct interaction between these proteins(Kamradt et al., 2001). Binding of the molecular chaperones αA- and αB-crystallin to the proapoptotic factors Bax and Bcl-Xs prevents their translocation into mitochondria, and could serve as an additional mechanism of apoptosis inhibition (Mao et al.,2004).

So far, all evidence that the small heat shock proteins αA- andαB-crystallin inhibit apoptosis are based on in vitro or tissue culture experiments. No animal models have yet been analyzed to elucidate the roles of these proteins in protecting cells from programmed cell death. In the present report, we describe a physiological role of α-crystallin in mammalian lens development, employing a knockout mouse model. We chose to eliminate bothαA- and αB-crystallin because they both might affect apoptosis,and absence of only one subunit could be partially or fully compensated by the remaining subunit. Our data suggest that the absence of α-crystallin causes elevated caspase activity in lens secondary fiber cells. As a result,lens secondary fiber cells fail to halt their apoptosis-like maturation program after degradation of the cell nuclei, mitochondria and all other organelles, and progress to the stage of cell disintegration.

Generation of αA-/αB-crystallin DKO mouse

The αA-/αB-crystallin DKO mouse line was produced by breedingαA-crystallin KO mice (Brady et al.,1997) to αB-crystallin KO mice(Brady et al., 2001). Both KO mice were generated and maintained in the 129Sv background. These mice also lack the HSPB2 gene product (Brady et al.,2001).

Preparation of the lens extract

The eyes were removed from mice immediately after euthanasia, and then further dissected with microsurgical scissors. Lenses were removed and cleaned from bound tissues. Dissected lenses, typically 8 to 10, were placed immediately into 200 μl of ice-cold caspase 3 assay buffer (20 mM HEPES, pH 7.4, 2 mM EDTA, 0.1% CHAPS, 5 mM DTT). Lenses were disrupted on ice in a 1.5 ml microtube, with a disposable polypropylene pestle rotated by a cordless drive unit. Homogenized extracts were centrifuged at 20,000 gfor 30 minutes at 0°C. After centrifugation, the supernatant was collected, aliquoted and stored at -80°C for future use. Protein concentration was measured with a Coomassie protein assay reagent kit(Pierce), which is based on the Bradford method(Bradford, 1976). A commercial solution of bovine serum albumin (BioRad) was used to prepare protein concentration standards from which standard curves were plotted.

Caspase 3 and caspase 6 activity assays

The colorimetric caspase 3 assay kit and fluorimetric caspase 6 assay kit(both from Sigma) were used in 96-well plate format, according to manufacturer's instructions. For each reaction, 160-230 μg of lens extract protein were used. Reactions usually were incubated 16-20 hours at room temperature, and the amount of cleaved product was measured on the microplate reader (Bio-Tek Instruments), or on a fluorometer (Perkin Elmer). Specific activity for caspase 3 was calculated in nmol of pNA/hr/ng of total protein,and for caspase 6 in nmol of AMC/hr/mg of total protein. Caspase 3 and caspase 6 activity data were analyzed using linear models in R (Version 2.1.1). In all cases, P≤0.007 and was highly significant.

Histology and immunofluorescence

For light microscopy, paraffin sections of mouse lenses at the different ages were stained with Hematoxylin and Eosin. Images were collected on a Zeiss Axioplan 2 photomicroscope with an attached CCD camera (OPELCO). For immunohistochemistry, frozen sections were used. A rabbit anti-active caspase 3 monoclonal antibody (BD PharMingen) was used undiluted, and a rabbit anti-caspase 6 polyclonal antibody (Sigma) was diluted 1:100 in 1×PBS. Freshly dissected eyes were embedded in OCT compound (Tissue-Tek, Miles Scientific) and 10-15 μm slices were cryosectioned at -20°C. Cryosections were fixed in acetone at -20°C for 5 minutes, washed three times in room temperature PBS, and blocked for 30 minutes with 10% normal goat serum (Vector Laboratories) in PBS at room temperature. Sections were incubated with primary antibody overnight at 4°C, then washed three times for 10 minutes with PBS. They were then incubated with Alexa Fluor 488-conjugated anti-rabbit IgG secondary antibodies (Molecular Probes),diluted 1:100 in 1× PBS, for 1 hour at room temperature, followed by a thorough washing (three times) with PBS; 0.1% Tween 20. For negative controls,primary antibodies were omitted. Distilled water was used for the final wash. Samples were mounted with aqueous mounting medium containing anti-fading agents (Biomeda) and examined with a Zeiss Axioplan 2 photomicroscope equipped with epifluorescence and a CCD camera (OPELCO). Fluorescence intensity was analyzed and quantified with Image-Pro Plus Scientific Software, version 5.1(Media Cybernetics). Fluorescence intensity value for anti-active caspase 3 and anti-caspase 6 staining was calculated as a difference between staining with and without primary antibodies.

TUNEL reaction

Paraffin-embedded eye sections were deparaffinized and rehydrated to PBS. Samples were treated with 10 μg/ml Proteinase K (ICN) in 10 mM Tris-HCl (pH 7.4) for 15 minutes at room temperature. Sections were washed three times in PBS solution after proteinase treatment. TUNEL labeling was performed with the In Situ Cell Death Detection Kit, Fluorescein (Roche) according to manufacturer recommendations (1 hour at 37°C). Fluorescein-labeled nucleotides, supplied with the kit, were used in the TUNEL reaction. For negative controls, terminal transferase was omitted. Sections were washed with PBS/0.1% Tween 20. If DAPI labeling was performed, sections were finally incubated with DAPI reagent (Molecular Probes) diluted 1:2500 in PBS for 15 minutes at room temperature and then washed with PBS/0.1% Tween 20. Distilled water was used for the final wash. Samples were mounted with aqueous mounting medium containing anti-fading agents (Biomeda) and examined with a Zeiss Axioplan 2 photomicroscope equipped with epifluorescence and a CCD camera(OPELCO).

Caspase 6 and α-crystallin pull-down interaction assay

We used the commercially available ProFound Pull-Down PolyHis Protein:Protein Interaction Kit (Pierce Biotechnology) to study possible direct protein-protein interactions. The assay was performed according to manufacturer's instructions. Purified recombinant C-terminal histidine tagged caspase 6 (Sigma) (7.2 μg in 400 μl of wash solution) was immobilized on the cobalt chelate gel as bait. In control experiments, caspase 6 was omitted. Lens extracts (68 μg), prepared as described above from 4.5-week-old wild-type mice, were used as the source of prey proteins, and were incubated in 400 μl of wash solution for 1 hour at room temperature with immobilized caspase 6. All wash and elution steps were performed as recommended in the assay manual.

Gel electrophoresis and western blot analysis

Proteins isolated from lens extract by the pull-down assay were analyzed by western blot. Proteins eluted from the cobalt chelate gel, ∼30 μl per lane, were electrophoresed on pre-cast 10% NuPAGE Bis-Tris gels with MES running buffer (Invitrogen). Proteins in the gel were then electroblotted onto 0.45 μm pore size nitrocellulose membranes (Schleicher & Schuell) for 120 minutes at 30 V in an XCell II Blot Module (Invitrogen). After transfer,membranes were blocked for 1 hour at room temperature with 0.2% I-Block reagent (Tropix) in PBS with 0.1% Tween 20, then probed with specific primary antibodies. The following primary antibodies were used in western blot analyses: rabbit anti-rhαA-crystallin polyclonal antibody (from Dr Joseph Horwitz, UCLA) diluted 1:5,000; rabbit anti-rhαB-crystallin polyclonal antibody (from Dr Joseph Horwitz, UCLA) diluted 1:5,000; and rabbit anti-caspase 6 polyclonal antibody (Sigma) diluted 1:2,500. Nitrocellulose membranes were incubated for 1 hour with primary antibodies, then washed three times for 5 minutes in PBS/0.1% Tween 20. Caspase 6 andαA/αB-crystallin were then visualized with the Western-Star(Tropix) chemiluminescent immunoblot detection system, according to the manufacturer's instructions.

Morphological changes in DKO mouse secondary lens fiber cells

Lenses from DKO mice (Fig. 1B-F) are appreciably smaller and softer than those from wild-type mice (Fig. 1A). The DKO lenses are opaque at birth, and exhibit a series of abnormal histological changes with age. By contrast, all lenses from wild-type animals were completely transparent.

Histological sections of a DKO mouse lens at 6 weeks of age shows major disruption of normal architecture only at the equator(Fig. 1B). This has been seen in other genetically altered mice, and may be due to physical forces exerted on a weakened structure by zonule fibers, which attach the lens to the musculature involved in visual focusing. No other morphological abnormalities are apparent. However, with age, the lenses of αA-/αB-crystallin DKO mice show cellular disintegration in the region of the secondary fiber cells (Fig. 1C-F). Lenses from wild-type (Fig. 1A) andαB-crystallin single KO mice (data not shown) never displayed this type of cell disintegration. However, αA-crystallin single KO mouse lenses display a similar pattern of cell death in much older animals (data not shown). This could be additional evidence for the compensatory activity of the two α-crystallin subunits, particularly in the lens, whereα-crystallin concentration is exceptionally high.

Fig. 1.

Lens fiber cells disintegrate in αA-/αB-crystallin double knockout mice. Fixed sections of the lens from a wild-type mouse at the age of 3 months (A), and from DKO mice at the ages of 6 weeks(B), 8 weeks (C), 3 months (D), 10 months (E) and 13 months (F) were stained with Hematoxylin and Eosin. eq,equatorial/bow region of lens; ln, lens nucleus; df, disintegrated fiber cells. Scale bar: 250 μm.

Fig. 1.

Lens fiber cells disintegrate in αA-/αB-crystallin double knockout mice. Fixed sections of the lens from a wild-type mouse at the age of 3 months (A), and from DKO mice at the ages of 6 weeks(B), 8 weeks (C), 3 months (D), 10 months (E) and 13 months (F) were stained with Hematoxylin and Eosin. eq,equatorial/bow region of lens; ln, lens nucleus; df, disintegrated fiber cells. Scale bar: 250 μm.

Formation of secondary lens fiber cells from epithelial cells occurs in the equatorial/bow region of the lens. Throughout life, layers of newly formed fiber cells cover the older cells. As a result, older secondary fiber cells are located closer to the center of the lens. For 8-week-, 3-month- and 10-month-old DKO mouse lenses (Fig. 1C-E), cell disintegration is more pronounced adjacent to the lens nucleus, the region with the oldest secondary fiber lens cells. The cell disintegration area progressively increased with animal age from 8 weeks to 13 months (Fig. 1C-F). Cellular disintegration was not seen in the lens nucleus at any age examined(Fig. 1). It has been suggested that lens primary fiber cells loose there nuclei and other organelles through a non-apoptotic pathway (Vrensen et al.,1991).

Without a firm foundation to build on, newly differentiating cells at the lens equator of older mice fail to form neatly arrayed fiber cells that elongate symmetrically to the anterior and posterior lens, but instead they form amorphous globular cells that migrate to the anterior part of the lens(Fig. 1C-F). Unlike normal lens fiber cells (Fig. 1A), these globular cells do not appear to follow the normal apoptosis-like differentiation pathway (Boyle et al.,2003). Eventually, the accumulated cell mass in the anterior lens pushes the hard lens core through the posterior lens capsule(Fig. 1F), obliterating any semblance of a lens.

Immunofluorescent examination of the lens section

The process of secondary lens fiber formation has been described as a case of an `attenuated form of cell death'(Dahm, 1999). Our observation of morphological abnormalities in DKO lenses raised two important questions:(1) do these abnormalities in the lenses of αA-/αB-crystallin DKO mice result from failure to arrest the apoptosis-like maturation process?; and(2) does secondary lens fiber cell disintegration result from increased caspase activity in αA-/αB-crystallin DKO mice? The answer to the second question could clarify the first and suggest a possible caspase-dependent apoptotic pathway for lens secondary fiber cell differentiation. Data from other laboratories suggested direct interaction with, and inhibition of, caspase 3 by αB-crystallin(Mao et al., 2001; Kamradt et al., 2001; Kamradt et al., 2002). Detection of the active form of caspase 3 in lenses of DKO mice was chosen as a starting point for this study. We used immunohistochemical labeling of the active form of caspase 3 in frozen eye sections from the 7-week-old DKO mice(Fig. 2). Almost no signal was detected when primary anti-active caspase 3 antibody was omitted and only secondary antibody was used (Fig. 2C). However, anti-active caspase 3 antibody gives a strong signal in the same area where secondary lens fiber cell disintegration is evident in older mice (Fig. 2A). Merging the bright-field image of the same section(Fig. 2B) with the immunofluorescence image (Fig. 2A) allowed us to see that the majority of active caspase 3 localized in the oldest secondary lens fiber cells(Fig. 2D). The presence of active caspase 3 in the area of imminent cellular disintegration, before cell disintegration becomes histologically apparent, suggests a caspase-dependent cell death pathway. In other words, the elevated level of caspase activity is probably causing the observed cell disintegration, which encouraged us to continue this line of research. The next logical step was to compare caspase activity in the lens from wild-type and αA-/αB-crystallin DKO mice.

Caspase activity in wild-type and αA/αB-crystallin DKO mouse lens extracts

For the initial screening of caspase activities in lens extracts, we used a wide spectrum of specific caspase inhibitors from R&D Systems: Z-VAD-FMK;Z-WEHD-FMK; Z-VDVAD-FMK; Z-DEVD-FMK; Z-YVAD-FMK; Z-VEID-FMK; Z-IETD-FMK;Z-LEHD-FMK; Z-AEVD-FMK; Z-LEED-FMK (data not shown). The most promising candidates for further detailed study, chosen from the preliminary screening,were caspase 3 and caspase 6. Both caspase 3 and caspase 6 belong to the subclass of executioner caspases. Interaction of α-crystallin and caspase 3 has already been described in tissue culture systems. However,almost nothing is known about the possible role of α-crystallin in regulation of caspase 6 activity or maturation.

Commercially available kits were used for the determination of caspase activity. Protease activity was assayed using chromogenic or fluorogenic sequence-specific substrates, DEVD-pNA for caspase 3 and VEID-AMC for caspase 6, in the presence or absence of specific inhibitors. This type of caspase activity determination has obvious flaws. Any protease present in the lens extract that can digest caspase 3 or caspase 6 substrate and can be blocked with the specific caspase inhibitor will be classified as caspase 3 or caspase 6. However, for easier reading, this paper will refer to DEVDase activity as caspase 3 activity and VEIDase activity as caspase 6 activity. Wild-type and DKO mouse lenses from three age groups [4-, 8- (or 9-) and 21- (or 23-)week-old] were chosen for extract preparation and protease assay. Caspase activities were assayed in wild-type and DKO lens extracts for each age group,in the presence or absence of a specific caspase inhibitor(Fig. 3). It is notable that,for all ages tested, DKO mouse lens extracts showed elevated activities of both caspase 3 (Fig. 3A) and caspase 6 (Fig. 3B) compared with wild type, with the DKO mice exhibiting two- to fourfold higher caspase 3 activity than wild type (Fig. 3A,C). Caspase 6 activity in DKO mouse lens extract was 1.3- to 3-fold higher than in wild-type depending on the age of the animal. Interestingly, both caspase 3 and caspase 6 activity levels from DKO mouse lenses demonstrate age dependency.

Fig. 2.

Immunostaining of DKO mouse lens with anti-active caspase 3 antibodies. Both primary anti-active caspase 3 antibody and secondary anti-rabbit IgG antibodies, conjugated with Alexa Fluor 488, were used for detection (A). The bright-field image was captured (B). As a negative control, the primary antibody was omitted, and only the secondary antibody was used (C). The bright field image (B) and active caspase 3 fluorescent image (A) were merged for morphological comparison (D). Scale bar: 250 μm; A-C are at the same magnification.

Fig. 2.

Immunostaining of DKO mouse lens with anti-active caspase 3 antibodies. Both primary anti-active caspase 3 antibody and secondary anti-rabbit IgG antibodies, conjugated with Alexa Fluor 488, were used for detection (A). The bright-field image was captured (B). As a negative control, the primary antibody was omitted, and only the secondary antibody was used (C). The bright field image (B) and active caspase 3 fluorescent image (A) were merged for morphological comparison (D). Scale bar: 250 μm; A-C are at the same magnification.

Age dependent caspase activity in wild-type andαA/αB-crystallin DKO mouse lens extracts

For clear illustration, caspase 3 and caspase 6 activities were plotted against animal age (Fig. 3C,D). Each activity value represents the difference in signal intensity between reactions with and without the specific caspase inhibitor. At all ages tested,both caspase 3 and caspase 6 activities were higher in extracts from DKO mice than from wild-type mice. The age dependence of caspase 3(Fig. 3C) and caspase 6(Fig. 3D) activities are remarkably similar. In lens extracts from wild-type mice, activity levels of both caspase 3 and caspase 6 remained relatively constant with age. However,caspase 3 and caspase 6 activity levels in the lens extract from DKO mice varied with the mouse age. In younger mice, both caspase activities follow a similar trend, very high at 4 weeks with a sharp decline by 8-9 weeks,suggesting that elevated caspase activity in the 4-week-old lenses is a harbinger of visible cell disintegration at a later time. We were able to see the first sign of cell disintegration histologically in 7- to 8-week-old lenses (Fig. 1C). By 8-9 weeks of age, most of the secondary lens fiber cells have disintegrated and are incapable of producing additional caspases, so there is a decrease in the overall caspase activity of the lens. Up to this point, morphological changes and caspase activities are entirely consistent. After 8 weeks, the caspase 6 activity remained low, increasing by only 15% at 23 weeks of age. However,caspase 3 activity in the lenses of 21-week-old mice returned to the same level of activity observed in 4-week-old mice. The increase in activity,particularly DEVDase activity, is perplexing. It may derive from the globular cells accumulating in the anterior lens. These cells are completely different from any other lens cell type and have not been well characterized.

Immunostaining of DKO and wild-type mouse lenses with anti-active caspase 3 and anti-caspase 6 antibodies

To strengthen the caspase activity data from wild-type and DKO mouse lens extracts, we employed immunofluorescence microscopy of frozen sections stained with anti-active caspase 3 or anti-caspase 6 antibodies(Fig. 4). Frozen sections from 7-week-old mouse eyes were labeled with antibody (described in detail in the Materials and methods section), and all images were captured with identical parameters. For fluorescent microscopy analysis, we selected an anterior portion of the lens, consisting of secondary fiber cells, and approximately in the middle of secondary fiber cell region, as illustrated on the lens diagram(Fig. 4A). The negative controls (Fig. 4D,E,H,I), in which primary antibodies were omitted, were carried out simultaneously with the experimental samples. All controls showed low background signal compared with sections where both primary and secondary antibodies were used(Fig. 4B,C,F,G). These immunofluorescence staining experiments for active caspase 3 and caspase 6, in a specific region of the lens, support our biochemical data demonstrating the presence of higher levels of caspase 3 and caspase 6 activities in DKO lenses. Secondary lens fiber cells from DKO mice showed a much more intense signal for both the active form of caspase 3 (Fig. 4C) and caspase 6 (Fig. 4G), than cells from the similar lens regions in wild-type mice(Fig. 4B,F).

Fig. 3.

Caspase 3 and caspase 6 activities. Caspase 3 (A,C) and caspase 6 (B,D) activities were measured in the presence (+) or absence(-) of a specific inhibitor (A,B) in lens extracts from wild-type and DKO mice from three age groups. The activities plotted in C and D as a function of animal age represents the difference in caspase activity measurements in the presence or absence of a specific inhibitor. Specific activities are in nmol of pNA/hr/ng of total extract protein for caspase 3 or in nmol of AMC/hr/mg of total extract protein for caspase 6. The error bars represent data error analyzed using linear models in R (Ver. 2.1.1). *P<0.001; **P=0.007; ***P=0.003.

Fig. 3.

Caspase 3 and caspase 6 activities. Caspase 3 (A,C) and caspase 6 (B,D) activities were measured in the presence (+) or absence(-) of a specific inhibitor (A,B) in lens extracts from wild-type and DKO mice from three age groups. The activities plotted in C and D as a function of animal age represents the difference in caspase activity measurements in the presence or absence of a specific inhibitor. Specific activities are in nmol of pNA/hr/ng of total extract protein for caspase 3 or in nmol of AMC/hr/mg of total extract protein for caspase 6. The error bars represent data error analyzed using linear models in R (Ver. 2.1.1). *P<0.001; **P=0.007; ***P=0.003.

Fig. 4

. Caspase 3 and caspase 6 immunostaining. Immunostaining of the anterior region of secondary lens fiber cells (A) in wild-type(B,D,F,H) and DKO (C,E,G,I) lenses with anti-active caspase 3(B,C) and anti-caspase 6 (F,G) antibodies. In negative controls (D,E,H,I),primary antibodies were omitted and only the secondary anti-rabbit IgG antibody, conjugated with Alexa Fluor 488, was used. (J) Image-Pro Plus v.5.1 software was used to quantitate fluorescence intensity in sections of lens from wild-type and DKO mice. Scale bar: 25 μm.

Fig. 4

. Caspase 3 and caspase 6 immunostaining. Immunostaining of the anterior region of secondary lens fiber cells (A) in wild-type(B,D,F,H) and DKO (C,E,G,I) lenses with anti-active caspase 3(B,C) and anti-caspase 6 (F,G) antibodies. In negative controls (D,E,H,I),primary antibodies were omitted and only the secondary anti-rabbit IgG antibody, conjugated with Alexa Fluor 488, was used. (J) Image-Pro Plus v.5.1 software was used to quantitate fluorescence intensity in sections of lens from wild-type and DKO mice. Scale bar: 25 μm.

Image-Pro Plus software was used to quantitate fluorescence intensities(Fig. 4J). For both caspase 3 and caspase 6, lens sections from DKO mice have more intense antibody staining compared with wild-type mice. Data for anti-active caspase 3 immunostaining,showing a 2.1-fold increase in DKO compared with wild type, are consistent with results obtained from enzymatic activity assays(Fig. 3A). However, in the case of caspase 6, we found a greater difference of 4.1 fold in staining intensity between sections from wild-type and DKO mice, compared with enzymatic activity data (Fig. 3B), which could be explained by the existence of degraded caspase 6. Immunoprecipitation experiments (data not shown) demonstrate the presence of low molecular weight polypeptides, in mouse lens extracts, that are recognized by anti-caspase 6 antibodies that probably are caspase 6 degradation products. The difference in degree of elevation of caspase 6 immunostaining and activity measurements in DKO mouse lenses could also be explained by the fact that immunostaining data reflect differences in the small specific subsection of the lens(Fig. 4A), while the enzyme activity experiments represent a difference in extracts from the entire lens.

TUNEL reaction

Fragmentation of nuclear DNA, which is often measured in situ by TUNEL reaction, is one of the hallmark characteristics of apoptotic cells. We employed TUNEL techniques to assess apoptosis in the bow regions of lenses from 7-week-old wild-type and 7.5-week-old DKO mice(Fig. 5). Most cells in the lens loose there nuclei, except for cuboidal epithelial cells at the lens anterior, which actively divide and form new layers of secondary lens fiber cells in equatorial/bow region. Compared with wild-type(Fig. 5A), lenses from DKO mice suffered severe disintegration in the bow region(Fig. 5B). We therefore selected for TUNEL analysis the area posterior to the bow region, which had remnants of the degenerated cells and their nuclei(Fig. 5B, boxed). Higher magnification of the selected regions of the bright field images(Fig. 5A,B, boxed) are shown in Fig. 5C,D. Contrary to the morphologically well organized fiber cells in lenses of wild-type mice(Fig. 5C), cells in different stages of disintegration are evident in lenses from DKO mice(Fig. 5D).

TUNEL analysis of lens sections from wild-type and DKO mice revealed a dramatic difference in DNA fragmentation. TUNEL-positive cells in lenses of wild-type mice (indicated by arrows in Fig. 5E) coincide with the narrow area of differentiating secondary fiber cells (merged images of bright field and TUNEL fluorescent signal are presented in Fig. 5G). Normally, these cells undergo an apoptosis-like process of maturation that includes DNA fragmentation and removal of nuclei. Compared with wild-type mice, significantly stronger TUNEL signal was observed in lenses from DKO mice(Fig. 5F). The large number of the TUNEL-positive cells in lenses from DKO mice coincides with the region of cell disintegration (indicated by arrows in Fig. 5H). Labeling conditions and imaging parameters were identical for wild-type and DKO samples. In both cases, wild-type (Fig. 5I) and DKO (Fig. 5J) lens sections demonstrate only background fluorescence when terminal transferase was omitted from the TUNEL reaction.

In order to determine the number of cells undergoing apoptosis in the lenses of wild-type and DKO mice, we combined TUNEL staining with DAPI, which labels nuclear DNA, on the same sections(Fig. 6). Similar areas of lens bow regions from wild-type and DKO mice were identically labeled and imaged. Cells in the peripheral zone of the lens bow region of wild-type mice have well stained nuclei (Fig. 6A)that gradually disappear in the area of secondary fiber cells maturation(Fig. 6G and marked with arrowheads in Fig. 6E). A majority of nuclei in lenses from wild-type mice was TUNEL negative(Fig. 6C). DNA fragmentation was detected in the narrow area of secondary fiber cell denucleation(Fig. 6C,E,G). Only one cell in this field from the thin area of secondary fiber cell maturation was double labeled with DAPI and TUNEL (arrow in the Fig. 6E). Other cells in this region, which exhibit punctuate TUNEL staining, have probably progressed to a late stage of DNA degradation with the nuclei completely disintegrated. In contrast to wild-type mice, the majority of nucleated fiber cells in lenses from DKO mice was TUNEL positive (Fig. 6B,D, see arrow in Fig. 6F), suggesting a high level of DNA fragmentation, which is one of the key characteristics of apoptosis. TUNEL-positive cells in lenses from DKO mice are coincident with the area of progressive cell disintegration and cell loss (Fig. 5H, Fig. 6H), suggesting the apoptotic character of this process. In negative control samples, no signal was detected above background fluorescence in wild-type(Fig. 6I) or DKO(Fig. 6J) lenses when terminal transferase was omitted in TUNEL reaction.

Fig. 5.

TUNEL labeling of lens secondary fiber cells. Bright-field image(A-D), TUNEL reaction fluorescence (E,F) and merge of both(G,H) sections from 7-week-old wild-type (A,C,E,G,I) and 7.5-week-old DKO (B,D,F,H,J) mouse lenses. In negative controls (I,J), calf thymus terminal deoxynucleotidyl transferase was omitted from the TUNEL reaction. C,D are enlargements of boxed areas in A,B. Arrows in E,F,H indicate TUNEL-positive nuclei. Scale bar: 125 μm in A,B; 25 μm in C-J.

Fig. 5.

TUNEL labeling of lens secondary fiber cells. Bright-field image(A-D), TUNEL reaction fluorescence (E,F) and merge of both(G,H) sections from 7-week-old wild-type (A,C,E,G,I) and 7.5-week-old DKO (B,D,F,H,J) mouse lenses. In negative controls (I,J), calf thymus terminal deoxynucleotidyl transferase was omitted from the TUNEL reaction. C,D are enlargements of boxed areas in A,B. Arrows in E,F,H indicate TUNEL-positive nuclei. Scale bar: 125 μm in A,B; 25 μm in C-J.

Fig. 6.

Secondary lens fiber cell apoptosis. DAPI labeling fluorescence(A,B), TUNEL reaction fluorescence (C,D), merge of both(E,F) and merge of DAPI and TUNEL with bright field (G,H) of lens sections from 7-week-old wild-type (A,C,E,G,I) and 7.5-week-old DKO(B,D,F,H,J) mice. Lenses on all images are oriented such that the upper right corner points towards the bow region and the lower left corner points towards the lens nucleus. In negative controls (I,J), terminal transferase was omitted from the TUNEL reaction. Arrows in E,F indicate nuclei double labeled with DAPI and TUNEL. Arrowheads in E indicate the area of secondary fiber cell maturation. Scale bar: 25 μm.

Fig. 6.

Secondary lens fiber cell apoptosis. DAPI labeling fluorescence(A,B), TUNEL reaction fluorescence (C,D), merge of both(E,F) and merge of DAPI and TUNEL with bright field (G,H) of lens sections from 7-week-old wild-type (A,C,E,G,I) and 7.5-week-old DKO(B,D,F,H,J) mice. Lenses on all images are oriented such that the upper right corner points towards the bow region and the lower left corner points towards the lens nucleus. In negative controls (I,J), terminal transferase was omitted from the TUNEL reaction. Arrows in E,F indicate nuclei double labeled with DAPI and TUNEL. Arrowheads in E indicate the area of secondary fiber cell maturation. Scale bar: 25 μm.

α-Crystallin interaction with caspase 6

Publications from other laboratories showed that caspase 3 co-immunoprecipitated with αB-crystallin(Kamradt et al., 2001; Mao et al., 2001), suggesting a direct protein-protein interaction. However, no data about caspase 6 interaction with α-crystallin has yet been reported. We used recombinant C-terminal histidine-tagged caspase 6 as bait immobilized on the cobalt chelate gel to study possible protein-protein interactions in wild-type lens extract. In order to evaluate any non-specific binding with the cobalt chelate gel, we prepared a control column from which caspase 6 was omitted. After extensive washing, proteins were eluted from the gel with a high concentration of imidazole and analyzed by western blot(Fig. 7). Detection with anti-caspase 6 antibodies (Fig. 7, upper panel) shows a strong caspase 6 signal. The major band of the eluted protein co-migrates with recombinant caspase 6 (left lane) applied onto the gel as control. The minor band below caspase 6 suggests a small degree of degradation during the pull-down assay. Surprisingly, we did not see any signal when we analyzed eluted fractions with anti-αB-crystallin antibodies (Fig. 7, middle panel), suggesting that, in contrast to caspase 3, caspase 6 has extremely weak or no interaction with αB-crystallin. Flow-through fractions from both control and caspase 6-bound columns analyzed with the same anti-αB-crystallin antibodies show very intense signal (data not shown),confirming the presence of αB-crystallin in the homogenate, and the ability of the antibody to detect it. Elution fractions were also analyzed by western blot with anti-αA-crystallin antibodies(Fig. 7, lower panel). A strong signal with anti-αA-crystallin antibodies in the elution fraction containing caspase 6 suggests a direct interaction of these two proteins. The anti-αA-crystallin antibodies used in this experiment exhibit no cross-reactivity with caspase 6 (Fig. 7, left lane). Minute amounts of αA-crystallin eluted from control Co2+ column (Fig. 7 middle lane) suggesting low level non-specific binding ofαA-crystallin to the cobalt chelate gel.

Fig. 7.

Western blots of cobalt chelate gel column elution fractions.Recombinant caspase 6 (0.6 μg, left lane), 30 μl of the elution fraction from the control column, where caspase 6 was omitted (middle lane), and 30μl of the elution fraction from the caspase 6-Co2+ column (right lane) were separated on a 10% NuPage gel and detected by western blot with anti-caspase 6, anti-αB-crystallin or anti-αA-crystallin antibodies.

Fig. 7.

Western blots of cobalt chelate gel column elution fractions.Recombinant caspase 6 (0.6 μg, left lane), 30 μl of the elution fraction from the control column, where caspase 6 was omitted (middle lane), and 30μl of the elution fraction from the caspase 6-Co2+ column (right lane) were separated on a 10% NuPage gel and detected by western blot with anti-caspase 6, anti-αB-crystallin or anti-αA-crystallin antibodies.

A role for α-crystallin in inhibition of caspase 3 activity was suggested from in vitro and tissue culture experiments. However, no animal model study has yet been used to probe the significance of α-crystallin as an anti-apoptotic agent. In this study, we demonstrated morphological abnormalities in lens secondary fiber cells of αA-/αB-crystallin DKO mice. In wild-type mice, where α-crystallin is a major structural protein in the lens, lens secondary fiber cells undergo a terminal differentiation process which parallels many aspects of apoptosis, but the process ceases before cell disintegration occurs. In the lenses of DKO mice,we hypothesize that these cells follow a more complete apoptosis-like pathway,resulting in cell disintegration. The fact that we were able to see initial cell disintegration in the older, more central layers of secondary fiber cells, is consistent with an apoptotic cell death pathway and demonstrates a possible role of the α-crystallin as an anti-apoptotic factor in this system. This notion is further supported by the immunohistochemical study with anti-active caspase 3 antibodies, in which the brightest signal for the active form of caspase 3 in the DKO mouse was located adjacent to the lens nucleus,the area of the oldest lens secondary fiber cells. This is exactly the region where we saw the initial cell disintegration in DKO mouse lens at the age of 8 weeks.

A caspase-dependent pathway of cell death was supported in our caspase activity experiments. Caspase 3 and caspase 6 activities are higher in the lenses of DKO mice than of wild-type mice under all conditions tested. Increased activities of apoptotic executioner caspases in these DKO lenses contribute to cell disintegration, which is consistent with α-crystallin inhibition of caspase activity in normal lenses, where α-crystallin is highly abundant. Interestingly, caspase 3 and 6 activity levels in wild-type mouse lenses remained relatively constant at all ages tested, and the lenses remained transparent throughout the life of the animal. However, activity levels of caspase 3 and caspase 6 in DKO mouse lenses fluctuated with the age of the animal, and changes in caspase activities were consistent with changes in lens morphology. Increased activities of apoptotic executioner caspases in the lenses of DKO mice at earlier ages contributed to the cell disintegration that was readily apparent at the age of 8 weeks. Partially disintegrated lenses have a decreased number of intact fiber cells capable of generating new caspase activity, which would lead to a decrease in the overall level of caspase activity in the lens. Increases in caspase 6 and caspase 3 activities at ages 23 and 21 weeks could reflect the impact of the other, non-fiber lens cells and result in further cell disintegration.

Comparing immunofluorescence signal intensities of similar regions in wild-type and DKO mouse lenses revealed an increase in caspase 6 and the active form of caspase 3 in lens secondary fiber regions in DKO mice. Quantitation of the fluorescent images show increased levels of both caspase 6 and the active form of caspase 3 in lens of DKO mice, compared with wild-type mice. Immunostaining of the active form of caspase 3 is consistent with the caspase 3 activity assay. The greater difference in staining intensity for caspase 6 in lenses from wild-type and DKO mice, compared with the data from caspase 6 activity analysis, can be explained by presence of caspase 6 degradation products in the lens. For both caspase 3 and caspase 6, elevated activities in this area are consistent with the site of future cell disintegration. Further characterization of caspases involved in terminal differentiation of lens secondary fiber cells is a crucial part of future studies.

As the majority of literature regarding the inhibition of apoptosis byα-crystallin has been focused on caspase 3, we chose this as our primary target. However, according to our preliminary caspase screening experiments with a wide spectrum of caspase inhibitors, caspase 6 plays a similar, or even more important, role than caspase 3 in secondary lens fiber cells maturation. A role for caspase 6 in lens secondary fiber cell maturation was suggested earlier (Wride et al., 1999; Foley et al., 2004), however,some data suggest that VEIDase activity in the lens is not attributed to caspase 6 (Zandy et al.,2005). Nevertheless, possible inhibition of caspase 6 byα-crystallin had not been previously described. Employing a pull-down assay we were able to demonstrate direct interaction between caspase 6 andαA-crystallin, which could provide a possible mechanism for modulation of caspase 6 activity. Interestingly, using the same assay, we did not find any interaction between caspase 6 and αB-crystallin. These data are consistent with our observation that lenses of αA-, but notαB-crystallin, single KO mice display secondary lens fiber cell disintegration, but only in much older animals. Perhaps αA-crystallin controls activity of caspase 6, which appears to be more important for lens fiber cell maturation, and αB-crystallin controls activity of caspase 3. Although interactions between αA-crystallin and caspase 3 have not yet been reported, we cannot rule out modulation of caspase 3 activity byαA-crystallin. This is an important goal of our ongoing study.

Employing a TUNEL technique, we visualized the precise narrow region in lenses from wild-type animals, where secondary fiber cells undergo terminal maturation and where denucleation occurs. In contrast to DKO animals,TUNEL-positive lens cells in wild-type mice were identified only in this area. TUNEL signal in lenses from DKO mice was more intense, suggesting a higher level of DNA fragmentation compared with samples from wild-type mice. In addition, almost every nucleated fiber cell in the lens of DKO mice,regardless of location in the lens, was TUNEL positive. Regions of morphological change, or cell loss, in lenses from DKO mice coincide with intense TUNEL signal, suggesting an apoptotic character of cell disintegration. A high level of endonucleolysis, and complete cleavage of nuclear DNA, is considered the key event of apoptosis. Elevated TUNEL signal in lenses from DKO mice suggests an inhibitory role of α-crystallin in apoptosis. This is consistent with the increased level of ischemia-induced apoptosis observed in the hearts of αB-crystallin/HSPB2 gene knockout mice compared with wild type (Morrison et al., 2004).

The mechanism by which αA- and αB-crystallin inhibit caspase activity is unknown. Data presented in this report and in publications from other laboratories suggest possible pathways of inhibition:αB-crystallin could interact with factors promoting apoptosis, e.g. Bax and Bcl-Xs (Mao et al., 2004),or the interaction could be directly between caspase 6 and αA-crystallin and between caspase 3 and αB-crystallin(Kamradt et al., 2001; Kamradt et al., 2002). If a direct protein-protein interaction is involved in regulation of caspase activity by α-crystallin, it could be at the stage of maturation of the executioner procaspase(s) to their fully processed, active form, as suggested earlier (Kamradt et al., 2001; Kamradt et al., 2002), or direct inhibition of the fully processed caspase, or both. Further elucidation of the mechanisms regulating caspase activities is essential.

This research was supported by the Intramural Research Program of the NIH,NEI. We thank Dr Terry A. Cox from Statistical Methods and Analysis Section,Biometry Branch, NEI, NIH for performing statistical analysis of caspase activity data; Dr Robert Fariss and Ms Amanda Bundek, both from NEI, NIH, for the confocal microscopy study not included in this report, which supports our immunofluorescent data and suggests elevated level of active caspase 3 in the lens of DKO mouse. We thank Dr Joseph Horwitz for providing antibodies toαA- and αB-crystallin; and Dr Steven Bassnett for helpful discussions and sharing his unpublished findings.

Andley, U. P., Song, Z., Wawrousek, E. F., Fleming, T. P. and Bassnett, S. (
2000
). Differential protective activity of alpha A- and alphaB-crystallin in lens epithelial cells.
J. Biol. Chem.
275
,
36823
-36831.
Aoyama, A., Frohli, E., Schafer, R. and Klemenz, R.(
1993
). Alpha B-crystallin expression in mouse NIH 3T3 fibroblasts: glucocorticoid responsiveness and involvement in thermal protection.
Mol. Cell. Biol.
13
,
1824
-1835.
Boyle, D. L., Takemoto, L., Brady, J. P. and Wawrousek, E. F. (
2003
). Morphological characterization of the AlphaA- and AlphaB-crystallin double knockout mouse lens.
BMC Ophthalmol.
3
,
1
-11.
Bradford, M. M. (
1976
). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding.
Anal. Biochem.
72
,
248
-254.
Brady, J. P., Garland, D., Duglas-Tabor, Y., Robison, W. G.,Groome, A. and Wawrousek, E. F. (
1997
). Targeted disruption of the mouse alphaA-crystallin gene induces cataract and cytoplasmic inclusion bodies containing the small heat shock protein alphaB-crystallin.
Proc. Natl. Acad. Sci. USA
94
,
884
-889.
Brady, J. P., Garland, D. L., Green, D. E., Tamm, E. R., Giblin,F. J. and Wawrousek, E. F. (
2001
). αB-crystallin is lens development and muscle integrity: A gene knockout approach.
Invest. Ophthalmol. Vis. Sci.
42
,
2924
-2034.
Dahm, R. (
1999
). Lens fiber cell differentiation - a link with apoptosis?
Ophthalmic Res.
31
,
163
-183.
Dasgupta, S., Hohman, T. C. and Carper, D.(
1992
). Hypertonic stress induces alpha B-crystallin expression.
Exp. Eye Res.
54
,
461
-470.
Foley, J. D., Rosenbaum, H. and Griep, A. E.(
2004
). Temporal regulation of VEID-7-amino-4-trifluoromethylcoumarin cleavage activity and caspase-6 correlates with organelle loss during lens development.
J. Biol. Chem.
279
,
32142
-32150.
Harding, J. J. (
1991
). In
Cataract. Biochemistry, Epidemiology and Pharmacology, 1st edn
, pp.
71
-81. London: Chapman and Hall.
Kamradt, M. C., Chen, F. and Cryns, V. L.(
2001
). The small heat shock protein alpha B-crystallin negatively regulates cytochrome c- and caspase-8-dependent activation of caspase-3 by inhibiting its autoproteolytic maturation.
J. Biol. Chem.
276
,
16059
-16063.
Kamradt, M. C., Chen, F., Sam, S. and Cryns, V. L.(
2002
). The small heat shock protein alpha B-crystallin negatively regulates apoptosis during myogenic differentiation by inhibiting caspase-3 activation.
J. Biol. Chem.
277
,
38731
-38736.
Li, D. W.-C., Xiang, H., Mao, Y.-W., Wang, J., Fass, U., Zhang,X.-Y. and Xu, C. (
2001
). Caspase-3 is actively involved in okadaic acid-induced lens epithelial cell apoptosis.
Exp. Cell Res.
266
,
279
-291.
Mao, Y.-W., Xiang, H., Wang, W., Korsmeyer, S. J., Reddan, J. and Li, D., W.-C. (
2001
). Human bcl-2 gene attenuates the ability of rabbit lens epithelial cells against H2O2-induced apoptosis through down-regulation of the alpha B-crystallin gene.
J. Biol. Chem.
278
,
43435
-43445.
Mao, Y. W., Liu, J. P., Xiang, H. and Li, D. W.(
2004
). Human alphaA- and alphaB-crystallins bind to Bax and Bcl-X(S) to sequester their translocation during staurosporine-induced apoptosis.
Cell Death Differ.
11
,
512
-526.
Mehlen, P., Preville, X., Chareyron, P., Briolay, J., Klemenz,R. and Arrigo, A. P. (
1995
). Constitutive expression of human hsp27, Drosophila hsp27, or human alpha B-crystallin confers resistance to TNF- and oxidative stress-induced cytotoxicity in stably transfected murine L929 fibroblasts.
J. Immunol.
154
,
363
-374.
Mehlen, P., Kretz-Remy, C., Preville, X. and Arrigo, A.-P.(
1996a
). Human hsp27, Drosophila hsp27 and human alphaB-crystallin expression-mediated increase in glutathione is essential for the protective activity of these proteins against TNFalpha-induced cell death.
EMBO J.
15
,
2695
-2706.
Mehlen, P., Schulze-Osthoff, K. and Arrigo, A.-P.(
1996b
). Small stress proteins as novel regulators of apoptosis. Heat shock protein 27 blocks Fas/APO-1- and staurosporine-induced cell death.
J. Biol. Chem.
271
,
16510
-16514.
Morrison, L. E., Whittaker, R. J., Klepper, R. E., Eric, F.,Wawrousek, E. F. and Glembotski, C. C. (
2004
). Roles for alpha B-crystallin and HSPB2 in protecting the myocardium from ischemia-reperfusion-induced damage in a KO mouse model.
Am. J. Physiol. Heart Circ. Physiol.
286
,
H847
-H855.
Piatigorsky, J. (
1981
). Lens differentiation in vertebrates. A review of cellular and molecular features.
Differentiation
19
,
134
-153.
Vrensen, G. F., Graw, J. and De Wolf, A.(
1991
). Nuclear breakdown during terminal differentiation of primary lens fibres in mice: A transmission electron microscopic study.
Exp. Eye Res.
52
,
647
-659.
Wistow, G. J. and Piatigorsky, J. (
1988
). Lens crystallins: The evolution and expression of proteins for a highly specialized tissue.
Annu. Rev. Biochem.
57
,
479
-504.
Wride, M. A. (
2000
). Minireview: Apoptosis as seen through a lens.
Apoptosis
5
,
203
-209.
Wride, M. A., Parker, E. and Sanders, E. J.(
1999
). Members of the Bcl-2 and caspase families regulate nuclear degeneration during chick lens fibre differentiation.
Dev. Biol.
213
,
142
-156.
Zandy, A. J., Lakhani, S., Zheng, T., Flavell, R. A. and Bassnett, S. (
2005
). Role of the executioner caspases during lens development.
J. Biol. Chem.
280
,
30263
-30272.