Contribution of Calpain Lp82–Induced Proteolysis to Experimental Cataractogenesis in Mice

Abstract

purpose. The purpose of the present experiments was to provide a biochemical mechanism for the involvement of lens-specific calpain Lp82 in experimental cataractogenesis in mice.

methods. Nuclear cataracts were produced by culturing lenses from 4-week-old mice and rats in calcium ionophore A23187 or by injection of buthionine sulfoximine (BSO) into 7-day-old mice. Casein zymography, sodium dodecyl sulfate–polyacrylamide gel electrophoresis, immunoblot analysis, calcium determinations, in vitro precipitation, and cleavage site analysis by mass spectrometry were performed on lens samples.

results. Amino acid sequences for Lp82 were found to be highly conserved in lenses from mouse to cow, and expressed Lp82 proteolytic activity was high in the mouse and rat. Lenses from mice were more susceptible to A23187-induced cataract and BSO cataracts than rats. Both types of cataracts showed rapid elevation of calcium, activation of Lp82 and m-calpain, and proteolysis of crystallins. Lp82 caused in vitro precipitation of crystallins; and in contrast to m-calpain, Lp82 truncated only the first five amino acids from the C-terminus ofα A-crystallin.

conclusions. Under pathologic conditions of massive elevation of lens calcium found in young rodent lenses, overactivation of Lp82 and m-calpain leads to rapid truncation of crystallins at both common and unique cleavage sites, precipitation of truncated crystallins, and cataract.

The ubiquitous m- and μ-calpains (EC 34.22.17) have been found in almost all animal cells, where they may have common regulatory functions. ¹ However, recent evidence suggests that the tissue-specific calpains (such as muscle p94, stomach nCL2, and nCL2′, digestive tract nCL4, and lens Lp82 and Lp85) may be more closely linked to tissue-specific functions. ² For example, muscle contains 2- to 10-fold more mRNA for muscle-specific calpain p94 and its isoforms than for ubiquitous calpains, ³ and mutations in p94 cause muscular dystrophy type 2A in humans. ⁴ Unexpectedly, the eye was found to contain tissue-specific variants of muscle p94. For example, Lp82 is a newly discovered lens-specific member of the AX1 subfamily of naturally occurring splice variants of muscle-type calpain p94. ⁵ mRNA levels for Lp82 in young rat lens were 3-fold higher than mRNA for p94 in muscle. ⁶ Lp82 and the other AX1 subfamily members show alternative exon 1 usage along with various combinations of splice variations. Members discovered so far include Lp82 and Lp85 in lens ⁵ and Rt88 in the retina. ⁷ This may indicate that common promoter elements in the eye cause the alternative transcription of AX1 from the p94 gene. The best characterized member, Lp82, shows high tissue specificity for lens, protein stability because of splice deletions in the IS1 and IS2 regions, high concentration in the insoluble fraction, and decreased sensitivity to the natural endogenous inhibitor of ubiquitous calpains, calpastatin. ⁸  

Cataracts are a leading cause of blindness throughout the world, and experimental cataracts in rats and mice have long been used to study basic biochemical mechanisms of cataract formation. Our previous finding of m-calpain cleavage sites on specific β-crystallins in lens strongly link overactivation of m-calpain to cataract formation in numerous rodent models. ⁹ However, most calpains have similar papain-like active sites, suggesting that Lp82-induced proteolysis could have been responsible for at least a portion of the proteolysis attributed to calpain. We recently discovered that Lp82 is the dominant isoform of calpain in young mouse lens. ¹⁰ Thus, the experiments reported below used two mouse models to assess the role of Lp82 induced–proteolysis in cataract formation. 

Methods

Caseinolytic Activity

Lenses without decapsulation were isolated from ICR mice (Charles River, Kanagawa, Japan), and Sprague–Dawley rats (Charles River) at 4 weeks of age. Experimental animals were handled in accordance with the Guiding Principles in the Care and Use of Animals (DHEW Publication, NIH 80-23) and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Lenses were first homogenized in buffer A containing 20 mM Tris (pH 7.5), 1 mM EDTA, 1 mM EGTA, and 2 mM dithioerythritol (DTE), and soluble proteins were obtained by centrifugation. ¹¹ Protein content was measured by the BCA Protein Assay (Pierce Chemical, Rockford, IL) following the recommendations of the manufacturer and using bovine serum albumin as standard. Caseinolytic activities were determined by casein zymography, using the method of Raser et al. ¹² Ten percent (1-mm thick) gels, copolymerized with 0.1% casein (TEFCO, Tokyo, Japan), were prerun with buffer containing 25 mM Tris (pH 8.3), 192 mM glycine, 1 mM EGTA, and 1 mM dithiothreitol (DTT) for 15 minutes at 4°C. Then each sample was loaded and run. After electrophoresis, the gels were incubated with slow shaking for 20 hours at room temperature in 20 mM Tris (pH 7.4), 10 mM DTT, and 2 mM calcium. Gels were stained with Coomassie brilliant blue. Bands of caseinolysis appeared white. 

Lens Culture

Lenses were obtained by a posterior approach from mice (ICR, Charles River) and rats (Sprague–Dawley, Charles River) at several ages. Lenses were cultured at 37°C under 5% CO₂ in 2 ml (mouse) or 4 ml (rat) Eagle’s minimum essential medium (MEM; GIBCO–BRL, Life Technologies, Rockville, MD) with 10% fetal bovine serum (normal group; GIBCO–BRL, Life Technologies). Ten micromoles of A23187 (Calbiochem, La Jolla, CA) was present on day 1 only (A23187 group). One hundred micromoles of E64d (Peptide Institute, Osaka, Japan) was present continuously in the A23187 + E64d control group. E64d, a cell permeable inhibitor of cysteine proteases such as calpain, was used because it had previously been shown to inhibit calcium ionophore cataract in cultured rat lenses. ¹³ Lenses were photographed under a dissecting microscope during the culture period. 

Calcium Determinations

The wet weight of lenses was measured, then dry weight was determined after heating at 100°C for 16 hours. Dried lenses were then digested in 0.2 ml concentrated HCl with gentle agitation at room temperature overnight, 0.8 ml water was added, and calcium content was measured by atomic absorption spectrophotometry (Polarized Zeeman Atomic Absorption Spectrophotometer model Z-8100; Hitachi, Tokyo, Japan). Lens water content was calculated by subtracting lens dry weight from wet weight, and calcium content in the lenses was then expressed in millimoles. 

Sodium Dodecyl Sulfate–Polyacrylamide Gel Electrophoresis and Immunoblot Analysis

After the culture, lenses were homogenized in buffer A, soluble and insoluble proteins were obtained by centrifugation, and protein concentrations were measured by the BCA assay. To assess proteolysis, sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) of lens proteins was performed on 12% gels. ¹⁴ Gels were stained with Coomassie brilliant blue R-250 and dried between cellophane sheets. Molecular masses were estimated by comparison to commercially available protein standards (broad range; Bio–Rad, Richmond, CA). Immunoblots for m-calpain, Lp82, and vimentin were performed by electrotransferring proteins from SDS-PAGE 12% gels onto polyvinylidene fluoride membrane (Millipore, Bedford, MA) using the method of Towbin et al. ¹⁵ Rabbit serum polyclonal antibodies against rat muscle m-calpain ¹⁶ and against a synthetic peptide from the Lp82 sequence ⁶ were used at 1:250 dilution. Mouse monoclonal antibody (ICN Pharmaceuticals, Costa Mesa, CA) against porcine lens vimentin purified from eye was used at 1:100 dilution. Immunoreactivity was visualized with alkaline phosphatase conjugated to anti-rabbit IgG or anti-mouse IgG secondary antibodies and 5-bromo-4-chloro-3-indolyl phosphate/nitroblue tetrazolium (BCIP/NBT; Bio–Rad). For measurement of endogenous calpain activity in lens soluble protein from cultured lenses, casein zymography was performed as noted above. 

Cataract Model In Vivo

Buthionine sulfoximine (BSO; Nacalai Tesque, Kyoto, Japan) was used to produce cataract in ICR mice. ¹⁷ Four subcutaneous injections of BSO (4 μM/g body weight) per day on postnatal days 7 and 8 were administered at intervals of 2.5 hours. Lenses were photographed under a dissecting microscope after careful dissection at each time period. Measurement of calcium contents, casein zymography, SDS-PAGE, and immunoblots was performed as described above. 

In Vitro Light Scattering

Lp82 was partially purified by 7.5 mm × 7.5 mm DEAE 5-pw column (TOSOH, Tokyo, Japan) chromatography from 12-day-old rat lens as previously described. ⁵ Twenty-seven microliters of Lp82 preparation was then incubated in 80 μl total soluble proteins (30 mg protein/ml) from 12-day-old rat lens with 1 mM Ca²⁺. Light scattering was measured as an increase in optical density at 405 nm using our previously published protocol, ¹⁸ except no KCl was added and endogenous calpains were inactivated by prior treatment of the total soluble lens proteins with 5 mM iodacetamide followed by excess DTE. 

Mass Spectrometry Analysis

Calpain isoform–specific cleavage sites were determined by incubating gel filtration–purified rat α-crystallins in a 50 μl mixture containing 10 to 20 μg α-crystallin, 20 mM imidazole (pH 6.8), 50 μM EGTA, 2 mM DTE, and partially purified Lp82 or rat recombinant m-calpain (1 μg/mg soluble protein; Calbiochem, San Diego, CA). Proteolysis was initiated by the addition of 1.6 to 3.0 mM CaCl₂, followed by incubation at 37°C for 30 minutes. Masses of C-terminally truncated α-crystallin were measured by electrospray ionization mass spectrometry using an ion trap mass spectrometer (model LCQ; Finnigan MAT, San Jose, CA). 

Statistical analysis of data was performed by Dunnett multiple comparison. 

Results

Cultured Lens Model of Cataract

Casein zymography revealed that along with m-calpain, Lp82 activity was high in 4-week-old mouse (Fig. 1A ) and rat (Fig. 1B) lenses. Relatively minor amounts of μ-calpain were also present. (Lp82 activity was also detected in domestic pig, guinea pig, rabbit, and bovine lenses but not in human lenses; data not shown.) Thus, the mouse and the rat with their fairly high concentrations of lens Lp82 were appropriate animals in which to test involvement of Lp82 in two models of cataractogenesis. 

Lenses from mice were found to be more susceptible to A23187-induced cataract than rats (Fig. 1 , middle). Addition of 10 μM A23187 during the first day of culture caused a dense opacity in the central region of the lens after only 2 days of culture. In contrast, lenses from 4-week-old rats required 4 days of culture in A23187 before a dense opacity appeared in the central region of the lens. Control lenses from 4-week-old mice and rats cultured in normal medium with no A23187 remained clear and appeared similar to fresh lenses. Calcium concentrations in mouse lenses cultured in A23187 increased faster than similarly cultured rat lenses (Fig. 1 , graph). Note that E64d inhibitors of cysteine proteases such as m-calpain and Lp82 prevented A23187 cataract in mouse lenses (Fig. 1 , middle). E64d did not prevent increases in lenticular calcium caused by A23187, indicating that the cataracts were dependent on calpain-induced proteolysis and not due to simple increases in calcium. 

Such massive increases in calcium activate calpains, and this is rapidly followed by degradation of these enzymes. ^{19 20} Loss of caseinolytic activity in the soluble proteins from lenses cultured in A23187 was therefore used as an indirect measure of calpain activation (Figs. 2A 2B ). At the beginning of the experiment, lenses from mice and rats both contained Lp82, μ-calpain, and m-calpain activities (lane 0). A23187-induced loss of all three enzyme activities occurred in mouse and rat lenses by day 2. A new broad band representing an active product from Lp82 ⁸ was also observed on day 2 in both species (Figs. 2A 2B ; hatched arrows). These changes in Lp82 and m-calpain were also confirmed by immunoblot analysis (data not shown). Proteolysis of lens crystallins accompanied activation of calpains and cataract formation in mouse and rat lenses incubated in A23187 (Figs. 2C 2D) . For example, βB1 polypeptide at 31 kDa and α-crystallin polypeptide at 20 kDa underwent limited proteolysis to form lower-molecular-weight fragments. This is a typical pattern resulting from proteolysis by calpain. ¹¹ Furthermore, the breakdown of vimentin, which is highly susceptible to proteolysis by m-calpain in lens, ²¹ was also observed in the soluble proteins from lenses cultured with A23187 on day 1 (Figs. 2E 2F) . 

In Vivo Cataract Model

Injection of BSO in mice to inhibit glutathione synthetase induced hazy opacity by day 3 and subsequent dense opacity by 5 days (Fig. 3A ). Lens calcium concentrations also increased on day 3 postinjection and continued to rise to very high levels on day 5 (Fig. 3B) . Use of the same BSO dosing in rats did not increase calcium or cause nuclear cataract formation. 

Three days after injection of BSO, cataractous mouse lenses exhibited losses in m-calpain, μ-calpain, and Lp82, along with the appearance of the broad band of partially degraded active Lp82 (Fig. 3C) , indicating calpain activation. Paradoxically, μ- and m-calpain activities appeared to recover by day 5. (Massive losses of extensively hydrolyzed calpains to the aqueous humor on day 5 may have left intact calpains in the lens to be revealed by zymographic analysis. Alternatively, growth of new fibers and calpain synthesis could have occurred.) Proteolysis accompanied calpain activation and BSO cataract formation. βB1 polypeptide at 31 kDa underwent limited proteolysis to form lower-molecular-weight fragments (Fig. 3D) , and vimentin was degraded (data not shown). 

In Vitro Light Scattering and Cleavage Site Analysis

The results found in the two mouse models of cataract noted above suggested that activation of Lp82 and m-calpain by elevated lenticular calcium caused proteolysis of crystallins and that then the partially truncated crystallins became insoluble and scattered light. In vitro incubation of rat lens crystallins with purified m-calpain is known to cause truncation of crystallins, insolubilization, and massive increases in light scattering. ²² To similarly test for in vitro light scattering by Lp82, Lp82 was first purified in the present experiments from larger rat lenses, as they have a more abundant source of Lp82 than the mouse. Enzymatically active, partially purified Lp82 (contains approximately 20% Lp85, an isoenzyme similar to Lp82 ⁵ ) from DEAE chromatography was then incubated with rat lens total soluble proteins. Lp82 caused a large increase in light scattering starting on day 5 (Fig. 4A ), which was associated with proteolysis of lens crystallins (Fig. 4B) . Light scattering was due to Lp82 proteolysis, because inhibition of Lp82 enzyme activity with E64 attenuated both light scattering and proteolysis (Fig. 4) . 

To test whether Lp82 and m-calpain degrade substrates differently, purified α-crystallin was incubated with each enzyme. Mass spectrometry analysis indicated that Lp82 caused a loss of the first five amino acids from the C-terminus of αA-crystallin (Fig. 5) . In contrast, incubation with m-calpain resulted in the removal of 10, 11, and 16 amino acids off the C-terminus of αA-crystallin. Similar mass spectrographic analysis results showed that both Lp82 and m-calpain cleaved the first 11 amino acids off the N-terminus of recombinant βA3-crystallin (data not shown). Thus, Lp82 and m-calpain produce common or unique cleavage sites depending on the crystallin substrate. 

Discussion

The experiments above described the role of a tissue-specific calpain, lens Lp82, in the universal pathologic response of lenses to most toxic and traumatic conditions—cataract formation. In a wide variety of young rat models of cataract, the early, common mechanism is believed to be truncation of crystallins by calpain activated by increased lens calcium. ⁹ This is followed by insolubilization of the truncated crystallins, light scatter, and opacity. In the present study, the mouse lens was found to be highly susceptible to A23187 cataract in culture and to BSO cataracts in vivo. Seven pieces of data described below supported the idea that Lp82, along with m-calpain, played a major role in the formation of these two cataracts. First, Lp82 caseinolytic activity was high in the 4-week-old mouse lens. This high level of Lp82 was observed in these adolescent mice even though Lp82 is known to decrease with lens maturation. ^{10 20} At the younger age of 12 days, Lp82 was the dominant calpain in mouse lens relative to m-calpain, ¹⁰ whereas in the present studies Lp82 activity was approximately the same as m-calpain. Second, calcium was markedly increased in both types of cataractous mouse lenses, and calcium levels in A23187 cataracts increased faster in mice than in rats. Domain IV of Lp82 contains protein sequence called EF hand structures potentially able to bind calcium. ²³ The calcium activation requirement of Lp82 for half-maximal activity in vitro (25 μM) is lower than reported for m-calpain (Shih M and Shearer TR, unpublished observations, November 1999). Thus, the extremely high lens calcium levels (>5 mM) observed in the two models were likely to activate both enzymes. Third, evidence for activation of Lp82 and m-calpain was present in both types of mouse cataracts as degradation of m-calpain, loss of Lp82, and formation of Lp82 with different mobility on native gels. Losses of Lp82 and m-calpain have also been observed in selenite cataract in the young rat. ²⁰ Autolytic degradation is a common feature after m-calpain activation, ¹⁹ but recent evidence indicates that purified Lp82 may be active without autolysis. 

Fourth, further evidence for involvement of Lp82 in mouse cataract formation was that proteolysis of crystallins occurred in both types of cataractous lenses. As previously observed, ¹⁰ the overall pattern of proteolysis of lens crystallins by Lp82 was similar in some respects to m-calpain, probably because both enzymes contain a papain-like cysteine catalytic site. The present studies more specifically reported that both Lp82 and m-calpain removed the first 11 amino acids from the N-terminal extension of βA3. Fifth, other evidence for involvement of Lp82 in mouse cataract formation was that E64d, a known inhibitor of Lp82 ⁸ and m-calpain, ¹³ ameliorated cataract formation and proteolysis in A23187 mouse cataract despite the fact that calcium levels were massively increased. This was similar to A23187 cataract in cultured rat lens ¹³ and indicated that activation of Lp82 and m-calpain and proteolysis of crystallins are intimately involved in mouse cataracts. We also recently showed that in vitro Lp82 is less sensitive to the endogenous calpain inhibitor calpastatin than m-calpain. ⁸ This weaker control of Lp82 is another reason why Lp82 is expected to be active in the formation of the two types of mouse cataracts in the present investigation. 

Sixth, our previous cleavage site studies with m-calpain suggested that truncation of rodent α- and β-crystallins alters normal protein–protein interactions, leading to precipitation. ^{24 25} In vitro precipitation and light scatter by young rat and mouse lens crystallins by m-calpain are well documented. ^{11 22} However, the data in the present study are the first to show that incubation of purified Lp82 with lens crystallins also causes in vitro precipitation. Seventh, furthermore, Lp82 produced a different cleavage site at five amino acids in the C-terminus of αA-crystallin between serine¹⁶⁸ and serine¹⁶⁹ compared with the m-calpain cleavage sites at −10, −11, and −16 amino acids from the C-terminus. The same Lp82-like serine¹⁶⁸–serine¹⁶⁹ cleavage site on αA-crystallin has been found in aged lenses from the cow and humans. ^{26 27} Because removal of C-terminal amino acids reduces the ability of α-crystallin to act as a molecular chaperone for damaged lens proteins, ^{24 28} the Lp82 cleavage site may be detrimental to the long-term stability of the lens during aging. Furthermore, the Lp82 cleavage site (ser¹⁶⁸–ser¹⁶⁹) may serve as a biochemical marker for Lp82 activity in various lenses. Although human lenses do not contain Lp82, the presence of the same cleavage site on αA-crystallin from human lenses suggests that a search for an Lp82-like protease may be fruitful. 

The present studies provided convincing evidence that Lp82 plays a role in cataract formation in young rodent models, yet mRNA levels and enzymatic activity for Lp82 decrease to very low levels by 3 months of age. ^{5 20} The major function of Lp82 under normal physiological conditions is therefore likely to be for lens development or cell remodeling during maturation of lens. With the exception of the human lens, Lp82 was highly conserved from mouse to cow. The cDNAs for Lp82 from lenses of mouse (GenBank accession No. AF091998), rat (U96367), rabbit (AF148956), domestic pig (AF148955), cow (AF148714), and humans were sequenced and compared. The same length for the open reading frame was present in all, except for Lp82 from human lenses. A deletion of four nucleotides produced a stop codon in exon 1 of the cDNA for human Lp82, and this was also observed in the human genomic sequence (Beckmann J, personal communication, January 1999). A comparison of the deduced amino acid sequences showed that Lp82 was highly conserved from mouse to cow. Based on the entire protein sequence, identities ranged from 94% to 99%. Even the most variable region, domain I derived from alternative exon 1, showed conservation between the species that ranged from 88% to 99%. Thus, Lp82-induced proteolysis may be a common event during normal lens development as well as in cataract formation in young rodent lenses. 

 Supported in part by NIH Grants EY05786 and EY03600 (TRS for both).

 Submitted for publication August 11, 1999; revised November 30, 1999; accepted December 30, 1999.

 Commercial relationships policy: N (MS, HM, LLD); E (YN, CF, MA); C5 (TRS).

 

Figure 1.

 

Left, Casein zymograms showing endogenous Lp82 (solid arrowhead), μ-calpain (open arrowhead), and m-calpain (hatched arrowhead) activities in lens soluble proteins (80 μg/lane) from 4-week-old mouse (A) and rat (B). Middle, Photomicrographs showing A23187 cataract in cultured lenses from 4-week-old mice and rats. Darker areas are nuclear cataracts in these backlit lenses. The graph shows total calcium concentrations in the lenses (mean ± SD, n = 4–5). No differences in lens calcium concentrations were observed between the A23187 group and the A23187 group + E64d group in either mice (present study) or rats. ¹³

Figure 2.

 

Casein zymograms showing activity of endogenous calpains in lens soluble proteins from 4-week-old mouse (A) and rat (B) lenses cultured in A23187 (80 μg/lane). SDS–PAGE of the lens soluble proteins from mouse (C) and rat (D) during formation of A23187-induced cataract (5 μg protein/lane). Molecular weight standards are indicated in kilodaltons on left (lane“ Marker”). Bands showing decreases are indicated with solid arrowheads, and new bands are indicated with hatched arrowheads. Immunoblots are shown for vimentin of lens soluble proteins from mouse (E) and rat (F) during formation of A23187-induced cataract (80 μg protein/lane). The lane marked Vimentin contains 0.2μ g purified vimentin. Bands showing decreases are indicated with solid arrowheads.

Figure 3.

 

(A) Photomicrographs of lenses dissected from mice developing BSO-induced cataract. Darker areas are cataracts in these backlit lenses. (B) Total calcium concentrations in mouse lenses with BSO cataracts (mean ± SD, n = 3–4, *P < 0.01 relative to normal containing 0.2 mM Ca²⁺). (C) Casein zymograms of endogenous calpain activities in the lens soluble proteins from BSO-treated mice (80 μg protein/lane). Solid arrows show intact Lp82 and m-calpain activities, and hatched arrow shows active product from Lp28. (D) SDS–PAGE of lens soluble proteins in BSO cataract (5 μg protein/lane). Bands showing decreases are indicated with solid arrows, and new bands are indicated with open arrowheads. Molecular weight standards are indicated in kilodaltons on left (lane M).

Figure 4.

 

(A) Increased light scattering, as measured by the increase in optical density at 405 nm, occurring during the in vitro incubation of partially purified Lp82 with rat lens soluble proteins (solid box datapoints). Solid diamond datapoints show attenuation of light scattering by inhibition of Lp82 with cysteine protease inhibitor E64. (B) SDS–PAGE of insoluble proteins from in vitro incubation in (A). Arrows indicate examples of proteins lost due to proteolysis in the Lp82 group not inhibited by E64. The first lane contains protein molecular weight standards indicated on the left in kilodaltons.

Figure 5.

 

Cleavage sites produced on αA-crystallin after in vitro incubation ofα -crystallin with Lp82 (stippled arrow) or m-calpain (solid arrows). The molecular masses (in kilodaltons) observed by mass spectrometry are compared with the calculated masses based on acetylated αA-crystallin for each peptide truncated from the C-terminus.