Molecular Vision 1998; 4:4 <http://www.emory.edu/molvis/v4/p4> Received 13 October 1997 | Accepted 10 February 1998 | Published 27 February 1998

Download Reprint

Cleavage of Beta Crystallins During Maturation of Bovine Lens

Marjorie Shih,¹ Kirsten J. Lampi,¹ Thomas R. Shearer,^1,2 and Larry L. David^1,2
 
 

¹Department of Oral Molecular Biology and ²Department of Ophthalmology, Oregon Health Sciences University, Portland, OR

Correspondence to: Larry L. David, Departments of Oral Molecular Biology and Ophthalmology, Oregon Health Sciences University, 611 S. W. Campus Drive, Portland, OR, 97201, email: davidl@ohsu.edu

Abstract

Purpose: (1) Identify major crystallin proteins in fetal and adult bovine lens, (2) examine the N-termini of ß-crystallins for truncation, and (3) determine if the protease m-calpain (EC 3.4.22.17) is responsible for the cleavage of bovine ß-crystallins during maturation.

Methods: Crystallins from fetal and adult bovine lenses were analyzed by one and two-dimensional electrophoresis and Edman sequencing of separated proteins and their tryptic fragments. Identical techniques were used to analyze crystallins following their incubation with purified m-calpain.

Results: The identities of the major crystallins and several additional crystallin species missing portions of their N-terminal extensions were identified in the fetal bovine lens. Besides the previously identified form of ßB1 missing 15 residues from its N-terminus, forms of ßA3 missing 11 and 22 residues were identified. With aging, the ßA3 (-22) species became a major protein in the adult bovine lens, and minor forms of ßB2 and ßB3 missing 8 and 22 residues from their N-termini, respectively, appeared. Purified m-calpain cleaved within the N-terminal extensions of bovine ß-crystallins and removed: 12 or 15 residues from ßB1; 8 residues from ßB2; 5 or 10 residues from ßB3; and 11 or 17 residues from ßA3.

Conclusions: Based on the cleavage sites in vitro, m-calpain may be partially responsible for cleavage of bovine ßB1, ßB2, and ßA3 during lens maturation. However, the preference of m-calpain to remove 12 residues from ßB1, and 11 and 17 residues from ßA3, suggested that the ßB1 (-15) and ßA3 (-22) species found in vivo were produced by a different protease. This unidentified protease may have a preference for the asparagine-proline-X-proline sequence found in the N-terminal extensions of ßB1 and ßA3.

Introduction

Bovine crystallins were some of the very first lens proteins whose sequences were determined [1]. As a result of these early studies, the modern nomenclature of the [alpha], ß, and [gamma]-crystallin subunits were partially defined based on the isoelectric points of these proteins. Each ß-crystallin subunit was named according to the order that individual ß-crystallin subunits emerged during ion exchange chromatography and their relative positions following isoelectric focusing [1]. Therefore, unlike the orthologous ß-crystallins found in other species, there was never a need to identify the major bovine ß-crystallin subunits on two-dimensional electrophoresis (2-DE) gels. Their identities were explicitly defined by their positions. However, many of the less abundant crystallin species observed on 2-DE gels of bovine lens crystallins remained unidentified due to post-translational modifications which altered both their relative molecular weights and isoelectric points [1]. The recent analysis of 2-DE separated crystallins from both human [2] and rat [3] lenses suggested that many of the post-translationally modified proteins in bovine lens could be ß-crystallins missing portions of their N-terminal extensions.

The loss of ß-crystallin N-terminal extensions is important because it may significantly alter crystallin interactions following lens maturation [3]. When the loss of N-terminal extensions is accelerated, increased light scatter occurs in experimental rodent models of cataracts [4]. The protease(s) responsible for the degradation of N-terminal extensions on ß-crystallin may differ between species. Results suggest the loss of ß-crystallin N-terminal extensions in rodent lenses is caused by the calcium-dependent protease m-calpain (also termed calpain II, EC 3.4.22.17) [3,5]. In contrast, the loss of ß-crystallin N-terminal extensions in human lenses may result from activation of an as yet unidentified protease(s) with a cleavage site specificity unlike that of m-calpain. These unidentified human lens protease activities may specifically recognize and cut between asparagine and proline within the asparagine-proline-X-proline (NPXP) sequence found in both human ßB1 and ßA3/A1 N-terminal extensions [2].

The purpose of this study was to first establish if ß-crystallin N-terminal extensions in bovine lens also undergo truncation during maturation, and then to determine if the cleavage sites in these crystallins were consistent with the activation of m-calpain, as found in rat lens, or activation of the putative NPXP recognizing protease, as observed in human lens. We found that ßB1, ßB2, ßB3, and ßA3/A1 all undergo partial cleavage in bovine lens, and that the observed cleavage sites in these proteins were consistent with the activation of both proteolytic activities. These data also demonstrate that the proteases responsible for loss of N-terminal extensions on ß-crystallins during lens maturation may differ between species. Maturationally related proteolysis in bovine lenses shares similarities to both rodent and human lenses.

Methods

Lenses were obtained by a posterior approach from fetal and adult bovine eyes (Ferry Brothers, Portland, OR) and homogenized in 20 mM imidazole (pH 6.8), 50 µM EGTA, 2 mM dithioerythritol, and 0.02% NaN₃. Soluble and insoluble proteins were obtained by centrifugation at 16,000 x g for 5 min and assayed for protein using the bicinchoninic acid (BCA) protein assay (Pierce, Rockford, IL) with bovine serum albumin as a standard. Lens proteins were separated by 2-DE using non-equilibrium pH gradient electrophoresis in the first dimension with pH 3.5-10 ampholytes, electroblotted onto polyvinylidene difluoride (PVDF) membranes, and subjected to direct N-terminal Edman sequence analysis as previously described [4]. Proteins containing blocked N-termini were identified by digestion of the 2-DE separated proteins on the surface of PVDF membranes with trypsin, separation of the resulting peptides by HPLC, and Edman sequencing of individual tryptic fragments [2,6]. It was not possible to calculate the pIs of the separated proteins, because they do not reach their isoelectric points in this non-equilibrium procedure. Equilibrium isoelectric focusing of crystallins using free carrier ampholytes is not advised, since the most basic crystallins are lost from the gel. The approximate molecular weights and pIs of the displayed regions of the gels containing the crystallins were determined by calculating the theoretical molecular weights and pIs of ßB1 (the highest molecular weight crystallin subunit with one of the most basic pIs) and [alpha]A-crystallin (the lowest molecular weight crystallin subunit with one of the most acidic pIs). Molecular weights and pIs of the N-acetylated forms of these proteins were calculated using the programs PAWS (Protein Analysis Worksheet version 8.1.1, by Dr. Ronald Beavis), and GeneWorks 2.5 (Oxford Molecular Group, Inc., Beaverton, OR), respectively.

The m-calpain cleavage sites within the N-terminal extensions of bovine ß-crystallins were determined by incubating soluble protein from fetal lenses with purified porcine heart m-calpain [7]. The use of porcine heart m-calpain was justified, since previous studies indicated that m-calpain isolated from different mammalian species and tissues exhibited similar cleavage site specificity [4,6]. The relative susceptibility of the various crystallin subunits to m-calpain digestion were estimated by incubating crystallins for a constant amount of time with increasingly higher concentrations of m-calpain. This resulted in more linear loss of substrate than did incubation with a constant amount of m-calpain for increasing lengths of time. This observation was likely due to autolytic inactivation of m-calpain in the presence of calcium [8]. The 0.05 ml m-calpain digestion mixture contained 250 µg of soluble protein from fetal bovine lens, 20 mM Tris (pH 7.5), 1.0 mM dithioerythritol, 1.25 mM CaCl₂, and a range of 0-1.25 units of m-calpain. One unit of m-calpain activity was defined as the amount of m-calpain producing 1 µg acid soluble fluorescein isothiocyanate labeled casein fragments per min at 30 °C [9]. Following incubation for 10 min at 30 °C, the reaction was stopped by addition of an excess of EGTA over CaCl₂. The N-terminal cleavage sites in the resulting partially degraded crystallins were determined by separating the proteins by 2-DE followed by direct N-terminal Edman sequence analysis as above.

One-dimensional and 2-dimensional electrophoretic gels prepared for image analysis or publication were stained with colloidal Coomassie Brilliant Blue G-250. Images were digitized using a Gel Doc 1000 camera (Bio-Rad, Hercules, CA), and analyzed with NIH Image (version 1.61, from the Research Services Branch, National Institutes of Health). The second dimension of the 2-DE gels was run on 1.5 mm thick, 16 x 14 cm, 12% polyacrylamide gels as previously described [4]. All 2-DE gels used in this study were loaded with 100-250 µg of protein. Single dimension SDS-PAGE gels were run on 1.5 mm thick, 8 x 8 cm, 18% polyacrylamide gels (Novex, San Diego, CA, catalog number EC6508) containing 5-6 µg of protein per lane.

Results

One dimensional electrophoretic analysis of fetal and adult bovine lens proteins

An overview of the pattern of proteins in fetal and adult bovine lenses by SDS-PAGE revealed that lens contained at least eight prominent bands ranging in molecular weight from approximately 22 to 32 kDa (Figure 1A). Adult lenses showed a less complicated pattern dominated by three to four bands at lower molecular weights ranging from 22 to 27 kDa. Analysis of 2-DE gels in experiments described below allowed identification of several of the changes observed in protein composition by one-dimensional electrophoresis shown in Figure 1. The protein bands corresponding to intact ßB1 and co-migrating ßA3/ßB3 decreased (marked in red), and two prominent protein bands corresponding to intact ßB2 and partially truncated ßA3 increased (marked in green) during aging. The protein pattern was similar between the soluble and insoluble fractions in both age groups, except that the adult insoluble fraction accumulated more low molecular weight components below 17 kDa (marked in purple) than the soluble fraction. These low molecular weight components may be similar to extensively degraded forms of crystallins described in human lenses [10-12]. This study focused instead on examining the larger, partially degraded crystallins in the range of 17-31 kDa. Since these partially degraded crystallins are best resolved using 2-DE, the proteins from both fetal and adult lenses were examined below using this technique.

Identification of crystallin subunits of fetal bovine lenses by 2-DE analysis and Edman sequencing

The major proteins of the water-soluble fraction of the fetal bovine lens were separated by 2-DE (Figure 2A) and individual species identified by partial Edman sequencing of either free N-termini, or internal tryptic fragments (Table I). Intact crystallin subunits in Figure 2 are labeled in black, while partially degraded crystallin subunits are labeled in red. The identities of the intact crystallin subunits labeled in Figure 2A and identified in Table I agreed with the previously published identities of soluble crystallins from calf lens cortex [1]. However, the present study identified several previously uncharacterized components produced by in vivo partial cleavage of ß-crystallin subunits. A previously observed component called ßB1b [1] was confirmed to contain a form of ßB1 missing 15 residues from its N-terminal extension [13]. However, unlike the previous study, an additional form of ßB1 missing 12 residues was not detected. Two partially degraded forms of ßA3 were found, missing either 11 or 22 residues from their N-terminal extensions. Since both ßA3 and ßA1 contain the same sequence past residue 17 of ßA3 due to the use of alternate in-frame initiation codons within a single ßA3/A1 m-RNA [14], it was unknown if the ßA3 (-22) species was produced by cleavage of either ßA3, ßA1, or both proteins. A lower molecular weight form of ßB3 was also observed in bovine lens (Figure 2A, asterisk). This partially degraded ß-crystallin was unusual, since it contained a blocked N-terminus, suggesting possible cleavage at its C-terminus. Both acidic and basic forms of [alpha]A and [alpha]B were identified in Figure 2A. The basic forms ([alpha]A2 and [alpha]B2) were likely the unmodified proteins, while the acidic forms ([alpha]A1 and [alpha]B1) were probably phosphorylated or deamidated forms [15]. Two regions containing [gamma]-crystallins were identified in Figure 2A. N-terminal sequence analysis of these unblocked proteins yielded identical sequences matching the reported residues 2-16 of both [gamma]D and [gamma]E. Therefore, it was unknown which of the two proteins were found in regions marked [gamma]D/E in Figure 2A. The crystallins in the water-insoluble fraction of fetal bovine lenses were also analyzed by 2-DE. This gel is not shown because it yielded a pattern which was identical to the 2-DE gel of soluble protein shown in Figure 2A.

Changes in crystallins during aging of bovine lens

During aging, bovine lens proteins undergo extensive alterations (Figure 2B). This resulted in a 2-DE pattern dominated by a few proteins and accumulation of degradation products (shown in red). The identities for ßB1, ßB1 (-15), ßB2, ßB3, ßA3, ßA4, and all [alpha]- and [gamma]-crystallins on the 2-DE gel of adult lens protein (Figure 2B) were inferred based on the positions of these proteins identified from fetal lenses (Figure 2A). The remaining labeled protein species in Figure 2B were identified by Edman sequencing. N-terminal sequence analysis of proteins below ßB2 identified four ß-crystallins missing portions of their N-terminal extensions. Two minor species of ßB2 and ßB3 missing 8 and 22 residues from their N-terminal extensions, respectively; and two major species of ßA3, both missing 22 residues from their N-terminal extensions, were identified (Figure 2B, Table II). The more basic ßA3 (-22) species on the left matched the position of the ßA3 (-22) species observed in fetal lenses (Figure 2A). The more acidic species on the right was likely due to deamidation of ßA3 (-22). The identity of the diffuse species directly above the acidic form of ßA3 (-22) and the species directly to its right remain unknown. As before, these partially degraded ßA3 crystallins could also have been produced by removal of 5 residues from the N-terminus of ßA1. The major protein species observed directly below the two ßA3 (-22) species was blocked at its N-terminus. However, this protein, which migrated to the same position as [gamma]S from fetal lens, was confirmed to be [gamma]S by the sequence of an internal tryptic peptide (Table II).

While losses of ßB1, ßB3, ßA3, and ßA4 occurred in adult bovine lenses, the concentration of intact ßB2, and partially truncated ßA3 (-22) increased dramatically (compare Figure 2A and Figure 2B). These changes are likely the result of both selective expression of ßB2 and ßA3/A1 crystallin genes, and selective degradation of ßA3/A1 crystallin protein. Similar 2-DE analysis of the water-insoluble fraction of adult bovine lens detected no difference in the protein composition between the water-soluble fraction shown in Figure 2B and water-insoluble fraction (gel not shown). This was unlike rat lens, where partially truncated ß-crystallin selectively entered the water-insoluble fraction [3].

Role of m-calpain in proteolytic changes in bovine lens

Since m-calpain activity has been demonstrated in bovine lens [16], and this calcium-dependent protease was previously implicated in the removal of N-terminal extensions in rat lens ß-crystallins [3], the susceptibility of bovine ß-crystallin to m-calpain digestion was determined. Water-soluble protein from fetal bovine lenses was incubated with either 0.25 or 1.25 units of purified m-calpain to determine the relative susceptibility of crystallins to the protease, and to detect possible intermediate products of degradation (Figure 3). The most abundant degradation product produced by m-calpain was in a region matching the position of a C-terminally degraded rat [alpha]B-crystallin produced by m-calpain (Figure 3, [alpha]B*) [17]. This was accompanied by the production of a species migrating to a similar position as C-terminally degraded rat [alpha]A-crystallin produced by m-calpain (Figure 3, [alpha]A*) [17]. The susceptibility of bovine [alpha]-crystallin C-termini to m-calpain digestion was previously reported [18].

Several ß-crystallins exhibited nearly equal susceptibility to m-calpain digestion as the [alpha]-crystallins. The ß-crystallins most susceptible to m-calpain digestion were ßA3 and ßB1. Incubation with 0.25 units and 1.25 units m-calpain caused the total loss of intact ßA3 and ßB1 spots (Figure 3, middle and bottom panels). Concurrent with these losses were the accumulation of degradation products of ßB1, ßB2, ßB3, and ßA3 missing various portions of their N-terminal extensions. These truncated ß-crystallins were identified by Edman sequence analysis which is summarized in Table III. The m-calpain degraded ßB1 by removing either 12 or 15 residues from the N-terminal extension. Removal of 12 residues from ßB1 by m-calpain was the preferred cleavage site, since only ßB1 (-12) appeared following incubation with 0.25 units of m-calpain (Figure 3, middle panel). In contrast, incubation with 1.25 units of m-calpain produced both ßB1 (-15) and ßB1 (-12) species which did not fully resolve from one another (Figure 3, bottom panel, and Table III). The production of additional ßB1 (-15) over that found before incubation can be most easily observed by comparing the top and bottom panels of Figure 3. Incubation with m-calpain also produced two cleavage products of ßA3 missing either 11 or 17 residues from their N-termini. The production of ßA3 (-11) was also difficult to observe following incubation with m-calpain, because like ßB1 (-15), this protein was already present before incubation. However, the mean density of the region containing ßA3 (-11) in the sample incubated with 0.25 units m-calpain increased by 27% over the mean density of ßA3 (-11) found in lens protein before m-calpain incubation. In contrast, the density of the ßA3 (-11) region in the sample incubated with 1.25 units m-calpain (Figure 3, bottom panel) was not significantly increased above the beginning density of the ßA3 (-11) region (Figure 3, top panel), because ßA3 (-11) underwent further cleavage to produce significant amounts of ßA3 (-17) in this sample. ßA3 (-17) was detected very near the ßA3 (-22) species observed before m-calpain incubation (Figure 3, bottom panel). For this reason, Edman sequence analysis of the region containing ßA3 (-17) also detected some ßA3 (-22) (Table III). However, unlike ßA3 (-17), there was little evidence that significant quantities of ßA3 (-22) were produced by m-calpain. Incubation with m-calpain also produced two forms of ßB3 missing either 5 or 10 residues from their N-terminus. While these two species were well resolved from one another, the ßB3 (-10) species migrated very near a cleavage product of ßB2 missing 8 residues from its N-terminus. This resulted in the simultaneous detection of both proteins during the Edman sequence analysis summarized in Table III. Note that the ßB3 (-10) cleavage product migrated to a position identical to the partially truncated form of ßB3 found before incubation which contained a blocked N-termini (Figure 3, top panel, ßB3*). It was unlikely that the ßB3 (-10) cleavage product was derived from this unusual form of ßB3, since the parent ßB3 diminished greatly during incubation with m-calpain. Furthermore, cleavage of ßB3* by m-calpain, even if it involved only removal of a N-terminal blocking group, would have likely altered the position of ßB3* on the 2-DE gel. Of interest was the finding that m-calpain cleaved rat ßB2, ßB3, and ßA3 at precisely the same positions within the N-terminal extensions as bovine ßB2 (-8), ßB3 (-10), and ßA3 (-11) when the N-terminal extensions of the three proteins from the two species were aligned [3].

Discussion

The present study confirmed the identities of the major crystallins of the fetal bovine lens, and detected several partially truncated forms of ß-crystallins not previously identified [1]. This additional mapping of modified proteins on 2-DE gels of adult bovine lens protein will also facilitate the future analysis of the other post-translational modifications of crystallins in bovine lens. The truncated form of ßA3/A1 crystallin found in fetal lens, ßA3 (-22), accumulated with age and became a major protein in the adult bovine lens. The abundance of the truncated ßA3 (-22) species on the 2-DE gels of adult bovine lenses was striking. The presence of both ßA3 (-22) and ßA3 (-11) in bovine lens was also recently reported by Werten et al. [19]. While the protease m-calpain could reproduce at least three of the cleavages in bovine ß-crystallins found in vivo, its apparent inability to reproduce the ßA3 (-22) cleavage suggested that at least one other proteolytic activity was also responsible for truncation of crystallins in adult bovine lens.

These results and those of earlier studies examining the maturational changes of crystallins in both rat and human lenses suggested that the loss of ß-crystallin N-terminal extensions occurs in many species. We hypothesize that this important event decreases the dispersive forces between crystallins and facilitates increases in protein concentration in maturing fiber cells. The higher protein concentrations and refractive index of the lens fiber cytosol may minimize the effect of protein aggregation. However, it remains largely speculative that this proteolysis is programmed and performs a specific function. Alternatively, it may just be a normal degradative process without function, amplified by the lack of protein turnover.

Determining which proteases remove the N-terminal extensions of ß-crystallins is important not only because it could provide information concerning normal lens development and function, but it will test the hypothesis that a loss of protease regulation during maturation could contribute to cataract. An accelerated loss of ß-crystallin N-terminal extensions in young rat lens during induction of cataracts by selenite administration was associated with rapid protein insolubilization and lens opacity [3]. Examination of truncated ß-crystallins in both normal and cataractous rat lens suggested that the protease m-calpain was responsible for the majority of ß-crystallin degradation in this species. In contrast, no evidence for degradation of human ß-crystallins by m-calpain in vivo was found. This conclusion is based on cleavage sites found in ß-crystallins from young human lenses [2], and m-calpain induced cleavage sites in human ß-crystallins (unpublished data). This suggested that an as yet unidentified proteolytic activity was responsible for ß-crystallin truncation in human lens. The present study suggested that the proteolysis of ß-crystallins in bovine lenses is performed by both m-calpain and additional unidentified protease(s) similar to that found in human lens.

The results suggesting that m-calpain was partially responsible for ß-crystallin truncation in bovine lenses is as follows. (1) Analysis of partially truncated ß-crystallins in both fetal and adult bovine lenses identified at least two truncated ß-crystallins containing known m-calpain cleavage sites. These were ßA3 (-11) found in fetal lenses, and ßB2 (-8) found in adult lenses. An earlier study also reported the presence of a partially truncated ßB1 in the cortex of calf lens missing 12 residues from its N-terminus [13]. While this species was not detected in the present study in vivo, m-calpain readily produced this cleavage in vitro (Table III). (2) The production of ßB1 (-15), the major cleavage product of ßB1 in bovine lens, may have also been caused by m-calpain. While ßB1 (-12) was the preferred m-calpain cleavage product of ßB1, incubation with a high concentration of m-calpain was capable of also producing ßB1 (-15). (3) While the other known m-calpain products ßB3 (-5), ßB3 (-10), and ßA3 (-17) were not detected in bovine lens, they may have been obscured in the 2-DE gels of adult lens protein by other modified crystallins, or they may have been further cleaved by other lens proteases. For example, the proposed m-calpain product ßA3 (-11) found in fetal bovine lens was absent in adult bovine lens. The m-calpain product ßA3 (-11) was likely cleaved by another protease to ßA3 (-22) during lens maturation. In a similar manner, other m-calpain products may have been further cleaved so they were no longer observed in mature lens.

The results suggesting that an additional as yet unidentified protease is also active in bovine lens is: (1) ßB1 (-15) and ßA3 (-22) found in bovine lens were also observed in human lens [2], a species where no evidence for proteolysis of crystallins by m-calpain has yet been found. (2) The cleavage sites within the N-terminal extensions of both ßB1 and ßA3 were within the sequence NPXP between asparagine and proline residues. The presence of proline residues at the new N-termini of both proteins could be interpreted as exopeptidase cleavage which paused at the prolyl residues. However, the presence of other prolyl residues, nearer the N-terminus of the intact crystallins than the prolyl residues at the cleavage site, and resistance of the other ß-crystallins to cleavage suggest a specific endopeptidase. This endopeptidase may recognize the NPXP site within the relatively long N-terminal extensions of ßB1 and ßA3. (3) The hypothetical protease cleaving within the NPXP sequence may have already been described in the lens, but its specific role in crystallin cleavage may not be recognized. Besides m-calpain, a partial list of endopeptidases demonstrated in lens include: several new proteases with activities against synthetic substrates [20], proteasome [21], serine protease [22], and membrane bound protease [23].

The present study illustrates the importance of interspecies comparisons while studying the mechanisms altering the structure of lens crystallins. Lenses from the longer lived cow may be a more appropriate model to understand the mechanism of crystallin modification in human lens than is the short lived rat. Maturation of crystallins in rat lenses may be predominately performed by m-calpain, while crystallin maturation in cow and human lens is more complex due to the involvement of other proteases and possible non-enzymatic hydrolysis. The bovine lens may be an appropriate tissue to purify and identify the proteolytic activities responsible for crystallin modification in human lens.

Acknowledgements

This study was partially funded by National Eye Institute grant numbers EY03600 (TRS) and EY07755 (LLD).

References

1. Berbers GA, Hoekman WA, Bloemendal H, de Jong WW, Kleinschmidt T, Braunitzer G. Homology between the primary structures of the major bovine beta-crystallin chains. Eur J Biochem 1984; 139:467-479.

2. Lampi KJ, Ma Z, Shih M, Shearer TR, Smith JB, Smith DL, David LL. Sequence analysis of betaA3, betaB3, and betaA4 crystallins completes the identification of the major proteins in young human lens. J Biol Chem 1997; 272:2268-2275.

3. David LL, Azuma M, Shearer TR. Cataract and the acceleration of calpain-induced beta-crystallin insolubilization occurring during normal maturation of rat lens. Invest Ophthalmol Vis Sci 1994; 35:785-793.

4. David LL, Shearer TR, Shih M. Sequence analysis of lens beta-crystallins suggests involvement of calpain in cataract formation. J Biol Chem 1993; 268:1937-1940.

5. David LL, Calvin HI, Fu SC. Buthionine sulfoximine induced cataracts in mice contain insolubilized crystallins with calpain II cleavage sites [letter]. Exp Eye Res 1994; 59:501-504.

6. David LL, Shearer TR. Beta-crystallins insolubilized by calpain II in vitro contain cleavage sites similar to beta-crystallins insolubilized during cataract. FEBS Lett 1993; 324:265-270.

7. David LL, Wright JW, Shearer TR. Calpain II induced insolubilization of lens beta-crystallin polypeptides may induce cataract. Biochim Biophys Acta 1992; 1139:210-216.

8. Hightower KR, David LL, Shearer TR. Regional distribution of free calcium in selenite cataract: relation to calpain II. Invest Ophthalmol Vis Sci 1987; 28:1702-1706.

9. David LL, Shearer TR. Purification of calpain II from rat lens and determination of endogenous substrates. Exp Eye Res 1986; 42:227-238.

10. Srivastava OP. Age-related increase in concentration and aggregation of degraded polypeptides in human lenses. Exp Eye Res 1988; 47:525-543.

11. Srivastava OP, McEntire JE, Srivastava K. Identification of a 9 kDa gamma-crystallin fragment in human lenses. Exp Eye Res 1992; 54:893-901.

12. Srivastava OP, Srivastava K, Silney C. Identification of origin of two polypeptides of 4 and 5 kD isolated from human lenses. Invest Ophthalmol Vis Sci 1994; 35:207-214.

13. Berbers GA, Hoekman WA, Bloemendal H, de Jong WW, Kleinschmidt T, Braunitzer G. Proline- and alanine-rich N-terminal extension of the basic bovine beta-crystallin B1 chains. FEBS Lett 1983; 161:225-229.

14. Hogg D, Tsui LC, Gorin M, Breitman ML. Characterization of the human beta-crystallin gene Hu beta A3/A1 reveals ancestral relationships among the beta gamma-crystallin superfamily. J Biol Chem 1986; 261:12420-12427.

15. Groenen PJ, Merck KB, de Jong WW, Bloemendal H. Structure and modifications of the junior chaperone alpha-crystallin. From lens transparency to molecular pathology. Eur J Biochem 1994; 225:1-19.

16. Yoshida H, Murachi T, Tsukahara I. Distribution of calpain I, calpain II, and calpastatin in bovine lens. Invest Ophthalmol Vis Sci 1985; 26:953-956.

17. Kelley MJ, David LL, Iwasaki N, Wright J, Shearer TR. alpha-Crystallin chaperone activity is reduced by calpain II in vitro and in selenite cataract. J Biol Chem 1993; 268:18844-18849.

18. Yoshida H, Yumoto N, Tsukahara I, Murachi T. The degradation of alpha-crystallin at its carboxyl-terminal portion by calpain in bovine lens. Invest Ophthalmol Vis Sci 1986; 27:1269-1273.

19. Werten PJ, Vos E, de Jong WW. Proteolysis of ßA3/A1 crystallin during normal aging of the bovine lens. Invest Ophthalmol Vis Sci 1997; 38:S297.

20. Sharma KK, Kester K, Elser N. Identification of new lens protease(s) using peptide substrates having in vivo cleavage sites. Biochem Biophys Res Commun 1996; 218:365-370.

21. Wagner BJ, Margolis JW. Age-dependent association of isolated bovine lens multicatalytic proteinase complex (proteasome) with heat-shock protein 90, an endogenous inhibitor. Arch Biochem Biophys 1995; 323:455-462.

22. Srivastava OP, Ortwerth BJ. Isolation and characterization of a 25K serine proteinase from bovine lens cortex. Exp Eye Res 1983; 37:597-612.

23. Srivastava OP, Srivastava K. Human lens membrane proteinase: purification and age-related distributional changes in the water-soluble and insoluble protein fractions. Exp Eye Res 1989; 48:161-175.

24. Kilby GW, Truscott RJ, Stuchbury GM, Sheil MM. Mass spectrometry of lens crystallins: bovine beta-crystallins. Rapid Commun Mass Spectrom 1996; 10:123-129.