Human and Monkey Lenses Cultured with Calcium Ionophore Form αB-Crystallin Lacking the C-Terminal Lysine, a Prominent Feature of Some Human Cataracts

Abstract

Purpose.: Elevation of lens calcium occurs in both human and experimental animal cataracts, and opacification may result from calcium-activated proteolysis. The purpose of the present study was to determine whether calcium accumulation in cultured human and Macaca mulatta lenses results in proteolysis of crystallins, the major lens proteins.

Methods.: Two-dimensional electrophoresis and mass spectrometry were used to construct detailed maps of human and monkey lens crystallins so that proteolysis after calcium accumulation could be monitored and the altered crystallins identified. Human and macaque lenses cultured in A23187 showed elevated lenticular calcium and superficial cortical opacities. The carboxypeptidase E (CPE) gene is expressed in human lens, and its presence in lens fibers was demonstrated by Western blot. To investigate whether CPE could cause similar truncation, purified αB-crystallin and CPE were incubated in vitro.

Results.: The major change observed in the crystallins of these cultured lenses was the accumulation of αB_1-174-crystallin resulting from the loss of a C-terminal lysine. This result was significant, because similar appearance of αB_1-174 is a prominent change in some human cataracts. αB-crystallin and CPE incubation result in the formation of αB_1-174-crystallin. This truncation was specific to αB_1-174-crystallin, since other crystallins were not proteolyzed. Although a weaker activator than zinc, calcium activated CPE in vitro.

Conclusions.: Since zinc concentrations did not increase during culture in A23187, calcium uptake in the lens may be responsible for CPE activation and αB_1-174 formation during cataract.

Cataracts are opacities in the lens of the eye. This disease is the leading cause of preventable blindness in the world, ¹ yet an effective therapeutic strategy to slow the rate of cataract progression is unavailable. The human lens faces several challenges in attempting to remain transparent during its long lifespan. It is exposed to several possible oxidants, ^2,3 undergoes a large number of posttranslational modifications, ⁴ and is unable to replace the damaged proteins contained in its oldest central region because of a programmed loss of organelles. ⁵ Opacification results from a loss of the short-range order of crystallins, the major proteins of the lens, due to their aggregation on unfolding or separation into protein-rich and -poor phases. ⁶ Three factors that may normally prevent opacification with increasing age are a healthy epithelium on the anterior surface of the lens where the bulk of metabolic activity resides, ⁷ a complex network of pumps and channels allowing movement of ions and small molecules to and from the lens interior, ⁸ and the maintenance of a reducing environment, especially in the center of the lens. ⁹ Although oxidation of proteins in the lens interior may directly precede opacification, alterations in lens epithelium may initiate the process. 

Global genomic ¹⁰ and proteomic ¹¹ analyses have been used to examine the differences between gene expression and protein composition in normal and cataractous human lenses. A major finding in the proteomic analysis, performed by two-dimensional gel electrophoresis (2-DE), was a 10% to 90% loss of the C-terminal Lys₁₇₅ from a major lens protein known as αB-crystallin. ¹¹ Additional changes in this normally 20-kDa protein also included an N-terminal cleavage yielding a 16.4-kDa fragment and an overall elevation in the concentration of αB-crystallin. 

αB-crystallin and its binding partner αA-crystallin belong to the small heat shock protein family and form a polydisperse aggregate of 300 to 1200 kDa. These proteins are capable of binding to partially unfolded intermediates of the β/γ-cystallin families and prevent their irreversible aggregation and precipitation during aging. ⁶ Thus, α-crystallins may function in the human lens, where they contribute 28% of the total protein, ¹² by preventing opacity, until they are “titrated out” by other unfolded lens proteins. ⁶ The study of lens epithelial cells from knockout mice missing α-crystallins suggests these proteins may also influence cell proliferation. Lens epithelial cells from αA^−/− mice exhibited reduced survival and growth, ¹³ whereas epithelial cells from αB^−/− mice exhibited genomic instability and hyperproliferation. ¹⁴ The reported cataract-specific changes in αB may alter one or more αB functions and contribute to cataract formation. 

The protease producing truncated αB-Lys_1-174 is unknown, but this cleavage was most likely due to the activation of a carboxypeptidase. Although aminopeptidases have been thoroughly examined in the lens, ¹⁵ carboxypeptidases have not been characterized. However, several metallocarboxypeptidases have been described in other tissues and are capable of removing basic residues from both peptides and proteins. Transcripts for carboxypeptidases E/H (NbLi0012; NEI Bank, available at http://www.neibank.nei.gov/, provided in the public domain by the National Eye Institute, Bethesda, MD), D, and X ¹⁶ have been detected in the human lens. Many human cataracts, especially those in the cortex contain increased levels of calcium, ¹⁷ and calcium is an activator of some proteases. Therefore, the purpose of the present study was to determine whether elevation of lens calcium in cultured primate lenses is capable of reproducing the previously reported alterations in lens αB-crystallin. In the present study, calcium loading in both human and monkey lenses caused the loss of the C-terminal Lys₁₇₅ from αB-crystallin, but did not produce the 16.4-kDa form of the protein. These results were unlike those in similar experiments with rodent lenses, where calcium loading caused activation of the calcium-dependent endopeptidase calpain. 

Materials and Methods

Source of Lenses

Human eyes from 3-, 71-, 76-, and 81-year-old donors were obtained from the Lions Eye Bank (Portland, OR). The research adhered to the tenets of the Declaration of Helsinki and was approved by the institutional human experimentation committee or institutional review board (IRB). Eyes from monkeys (Macaca mulatta) ranging in age from 1 to 12 years were obtained from procedures unrelated to the present studies. Experimental animals were handled in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and the NIH Guiding Principles in the Care and Use of Animals. 

Protein Preparation

Samples for 2-DE maps for soluble proteins from human lenses were prepared by homogenizing whole decapsulated lenses from a 3-year-old human donor in 1 mL of 20 mM sodium phosphate, 1.0 mM EDTA (pH 7.0) buffer, and centrifugation at 20,000g for 30 minutes. Soluble proteins from 4-year-old monkey lenses were similarly prepared, except the cortical and nuclear regions of the lenses were first isolated by dissection so that each contained approximately 50% of the lens fiber mass. For the analysis of the protein alterations caused by the accumulation of calcium, the lenses were washed with PBS three times after incubation. To obtain the epithelial samples, we carefully dissected the lens capsule with the epithelial cells from the lens attached and homogenized it in buffer A containing 20 mM Tris (pH 7.5), 1 mM EGTA, 1 mM EDTA, and 2 mM dithioerythritol (DTE). Soluble protein was obtained by centrifugation at 13,000g for 15 minutes at 4°C. To obtain concentric samples from the outer cortex of cultured lenses, each decapsulated lens was gently agitated for up to 2 hours in 1 mL buffer A in a 15 mL centrifuge tube. One milliliter of buffer A was collected every 5 minutes and replaced with 1 mL of fresh buffer A. Protein concentrations were measured in each fraction, and the samples were pooled to obtain 5 mg of protein designated as the superficial region of the lens cortex. Protein concentrations were measured by the protein assay (BCA Protein; Pierce Chemical Co., Rockford, IL) using BSA as a standard, and 400-μg portions of protein were dried by vacuum centrifugation in preparation for electrophoresis. 

Lens Culture

Lenses from human and monkey donors of several ages were obtained by a posterior approach. For human lenses, culture was performed within 48 hours after death and within 2 hours for monkeys. The lenses were dissected in Eagle's minimum essential medium (EMEM; ATCC, Manassas, VA) containing 200 U/mL penicillin and 200 μg/mL streptomycin. For a 30-minute preincubation, the dissected lenses were placed in 2 mL EMEM containing 100 U/mL penicillin, 100 μg/mL streptomycin, 0.25 μg/mL amphotericin, 50 μg/mL gentamicin, and 10% fetal bovine serum (Invitrogen Inc., Carlsbad, CA) using six-well cell culture plates. The lenses were then transferred to 2 mL EMEM with 50 μg/mL gentamicin and 10% fetal bovine serum and cultured at 37°C under 5% CO₂ for up to 7 days. Ten micromoles of A23187 (EMD Biosciences, Inc., San Diego, CA) were added on day 1 only (A23187 group). Paired lenses were used for control and experimental protocols. 

Two-Dimensional Gel Electrophoresis

First-dimension isoelectric focusing was performed using custom, 18-cm, linear pH 5 to 10 immobilized pH gradient (IPG) strips produced as previously described, ¹⁸ and reswelled in 340 μL of solution containing 400 μg protein, 8 M deionized urea, 2% CHAPS, 50 mM DTT, 2% glycerol, 2% pH 6 to 11 IPG buffer (GE Health Care, Piscataway, NJ), and a trace of bromophenol blue. Isoelectric focusing was performed (Protean IEF Cell; Bio-Rad, Hercules, CA) using a program with the voltage ramp set at rapid, a final 3,500-V setting, 50,000 total volt hour, 50 μA limit per gel, and 20°C temperature. The gel strips were then reduced/alkylated, and the second-dimension separation was performed on 24 × 18.5-cm 12% SDS PAGE gels, as previously described. ¹⁸ The gels were then either stained with Coomassie blue G-250 ¹⁹ in preparation for in-gel digestion and identification of proteins or negative stained with zinc-imidazole ²⁰ in preparation for protein elution and mass measurement. 

Protein Analysis by Mass Spectrometry

In-gel trypsinization was performed as previously described, ²¹ and digests were analyzed by either LC-MS ²¹ or atmospheric pressure MALDI-MS ²² to collect MS/MS spectra using an ion trap mass spectrometer (LCQ Classic; ThermoFinnigan, San Jose, CA). Proteins were identified with the system software (SEQUEST software; ThermoFinnigan) to correlate experimental MS/MS spectra with theoretical MS/MS spectra calculated from peptide sequences in a human subset of the Swiss Prot protein database (http://www.expasy.org/ provide in the public domain by the Swiss Institute of Bioinformatics, Geneva, Switzerland). The results were filtered using DTASelect. ²³ ΔCN values greater than 0.08 were required, and identified peptides had to exceed XCorr values of 2.0, 2.5, and 3.8 for +1, +2, and +3 ions, respectively, for the LC experiments using electrospray ionization; whereas +1 peptides derived by MALDI had to exceed an Xcorr value of 1.5. All identified proteins also required two or more unique peptide matches per entry for a positive identification. 

Mass measurement of proteins eluted from 2-DE gels was performed as previously described, ²⁴ with the following modifications. After negative staining with zinc-imidazole, spots from single gels were excised, shaken twice for 15 minutes in 192 mM glycine, 25 mM Tris base, 50 mM DTT, 0.1% SDS, and crushed by passing through a 20-μm stainless steel frit with a 0.5-mL gas-tight syringe. One-hundred fifty microliters 96 mM glycine, 12.5 mM Tris base, and 50 mM DTT was then added to the syringe to transfer the remaining gel particles into a centrifuge tube, and the resulting slurry was shaken for 30 minutes. The slurry was then transferred to a microcentrifuge filter (Ultrafree-MC, UFC30HV00; Millipore, Bedford, MA) and centrifuged for 15 minutes at 13,000g. An additional 50 μL of the above solution was added, and the device was centrifuged again. The collected liquid was then dried by vacuum centrifugation and redissolved in 50 μL of 5% formic acid. The masses of the eluted proteins were determined by injecting the sample onto a 1.0 × 250-mm C4 column. The same trap cartridge, column, and electrospray ionization technique was used as before, ²⁴ except a 20 μL/min flow rate and 2% to 60% acetonitrile gradient over 50 minutes was used, and 0.05% TFA was added to the mobile phase to prevent formation of SDS-protein adducts during mass analysis. 

SDS-PAGE and Immunoblot Analysis

SDS-PAGE of lens proteins was performed with 10% bis-Tris gels in MES buffer (NuPAGE; Invitrogen Inc.) for CPE detection. Immunoblot analysis was performed by electrotransferring proteins from the gels onto polyvinylidene fluoride (PVDF) membranes (Millipore, Bedford, MA) at 100 V (constant) for 60 minutes at ice-cold temperature using Tris-glycine buffer. Antibody source and dilution were goat polyclonal antibodies against CPE (Santa Cruz Biotechnology Inc., Santa Cruz, CA) used at 1:100. Immunoreactivity was visualized with alkaline phosphatase conjugated to anti-rabbit IgG secondary antibody and BCIP/NBT (Bio-Rad). 

CPE Assay

CPE was inactivated by incubating recombinant human (rh)CPE (>95% purity; R&D Systems, Inc., Minneapolis, MN) with 2 mM EGTA (Sigma-Aldrich, St. Louis, MO) in assay buffer containing 50 mM sodium acetate (pH 5.5) for 10 minutes at room temperature. After the ions were chelated, 50 μg/mL of CPE and 100 μg/mL of crystallins: human recombinant αB (Abcam, Inc., Cambridge, MA), human recombinant βB1 (kindly provided by Kirsten Lampi, Oregon Health & Science University, Portland, OR), and human recombinant γD/γS containing N-terminal His tags (kindly provided by Johnathan King, Massachusetts Institute of Technology, Cambridge, MA) were incubated with calcium or zinc for 1 hour at room temperature. The CPE inhibitor 1,10-penanthroline (Sigma-Aldrich) was used to confirm CPE-dependent αB degradation. Crystallin masses were measured after incubation by mass spectrometry, as described for proteins eluted from 2-DE gels. 

Zinc Measurement

Wet weight of the lenses was determined by weighing the lenses before and after heating at 80°C for 16 hours. For zinc measurement, each dried lens was digested overnight with gentle agitation in 1.0 mL concentrated HCl. Water was added to dilute the samples at 1:100 for lens, 1:2 for EMEM medium, and 1:5 for CPE assay samples, and zinc was determined at 206.20 nm by inductively coupled plasma emission spectrometry (ICP-ES; Optima 2000 DV; Perkin-Elmer Life and Analytical Sciences Inc., Boston, MA). Zinc content in the lenses was expressed as milliequivalents zinc per kilogram lens water. 

Results

Two-Dimensional Electrophoresis Maps of Human and Monkey Lens Proteins

Before analysis of the protein alterations caused by the accumulation of calcium in primate lenses, all major proteins resolved by 2-DE of soluble proteins from a noncultured 3-year-old human whole lens (Figs. 1A, 1B), and the cortex of a 4-year-old monkey lens (Figs. 1C and 1D) were trypsinized, subjected to mass spectral analysis, and identified (Tables 1, 2). The pattern found in the 3-year-old whole human lens was similar to that in a previous IPG-based, 2-DE gel of a human adult lens cortex. ²⁵ The similarity was due to the young age of the crystallins in the fibers of both samples. ²⁵ Comparison of human and monkey lens crystallins by 2-DE suggests that young lens fibers of both species contain similar high abundances of βB2-, αA-, αB-, βA3/A1-, and βB1-crystallin. Furthermore, most of the distinct crystallins in both species contain multiple forms due to similar partial truncation and heterogeneity of isoelectric points. This observation suggests that crystallins from monkey lenses are subjected to the same age-related truncation, deamidation, and phosphorylation as are human lenses. ^26,27 Since calcium elevation in rodent lenses causes rapid truncation of crystallins due to activation of proteases, ^24,28,29 experiments were next performed with cultured human and monkey lenses to determine whether calcium elevation could further enhance the observed truncation of crystallins. 

Figure 1.

 

2-DE gels of soluble proteins from 3-year-old whole human lens (A), and 4-year-old monkey lens cortex (C). Each gel image is duplicated in (B) and (D) to show regions numbered, excised, digested with trypsin, and proteins identified by mass spectral analysis (Tables 1, 2; Supplementary Data S1, S2).

Table 1.

 

Identification of Major Proteins from a 3-Year-Old Whole Human Lens

Spot Number	Protein	Accession No.
43, 46, 48, 49, 50, 53, 54, 55	αA	P02489
29, 32*, 36	αB	P02511
1–12, 16, 21, 24, 25	βB1	P53674
22, 23, 26, 27, 28, 34, 37	βB2	P43320
40*	βB3	P26998
31, 35, 41, 44, 51	βA3/A1	P05813
39, 42, 45, 47, 52	βA4	P53673
14, 17, 18, 32*, 33	γC	P07315
19, 20, 30	γD	P07320
38	γS	P22914
40*	HSP27	P04792
13, 15, 56, 57	NI	—

 From spots numbered, excised, digested, and identified from tandem MS spectra of proteins shown in Figure 1B. NI, not identified.

*  Denotes spots that contained more than one detected protein. Peptides identified from tandem MS data are summarized in Supplementary Data S1.

Table 2.

 

Identification of Major Proteins from the Cortex of a 4-Year-Old Monkey Lens

Spot Number	Protein	Accession No.
1, 2, 3*	Aldehyde dehydrogenase	P00352
4	Alpha enolase	P06733
10	Glyceraldehyde-3-phosphate dehydrogenase	P04406
14	Phosphatidylethanolamine-binding-protein	P30086
58	Fatty acid-binding protein	Q01469
59*	Actin	(P60709)
73*, 75*	HSP 27	(P04792)
3*, 56, 59*, 60, 61, 62, 64, 65, 68	αA	(P02489)
27, 45, 55, 69	αB	(P02511)
12, 13, 35, 36, 38, 39, 50	βB1	(P53674)
5, 20, 21, 32, 33, 34, 37, 40, 41, 47, 48, 49*, 52*, 53, 57, 72*	βB2	(P43320)
49*, 52*	βB3	(P26998)
30, 31, 42, 43, 46, 54*, 70, 72*, 73*, 74, 75*	βA3/A1	(P05813)
71, 76, 77	βA4	(P53673)
17, 18, 19	γC	(P07315)
ND	γD
23, 29, 54*	γS	(P22914)
6, 7, 8, 9, 11, 15, 16, 22, 24, 25, 26, 28, 44, 51, 63, 66, 67, 78, 79	NI

 Data are from spots numbered, excised, digested, and identified from tandem MS spectra of proteins shown in Figure 1D. ND, not detected.

*  Denotes spots that contained more than one detected protein. Peptides identified from tandem MS data are summarized in Supplementary Data S2.

Human Lens Culture in Calcium Ionophore

Culture of both human and monkey lenses in the calcium ionophore A23187 induced superficial cortical opacities. ³⁰ The proteins from these affected regions composed of the lens epithelium and superficial cortex were examined by 2-DE. In the absence of the calcium ionophore, the crystallins from the epithelium (Fig. 2A, normal) and superficial cortex (Fig. 2B, normal) from the lenses of a cultured 81-year-old human donor were remarkably similar to the crystallins from whole lenses of the 3-year-old donor (Fig. 1A). Culture in the absence of the calcium ionophore also did not induce changes in the lens crystallins; a 2-DE gel of epithelial protein from an uncultured 71-year-old donor lens was identical with lenses cultured for 7 days without ionophore (data not shown). 

Figure 2.

 

2-DE gels of the soluble proteins from the epithelium (A) and superficial cortex (B) of human lenses cultured without (Normal) or with A23187. Only the area containing the majority of crystallins is shown. Landmark spots for intact αA-, intact αB_1-175-, and βB2-crystallin are indicated in relation to the increased amounts of truncated αB_1-174 (arrow) observed only in A23187-treated lenses. (C) Note the similarity of the above data to the abundant truncated αB_1-174 found in cataractous human lenses. Adapted from Jimenez-Asensio J, Colvis CM, Kowalak JA, et al. An atypical form of αB-crystallin is present in high concentration in some human cataractous lenses. Identification and characterization of aberrant N- and C-terminal processing. J Biol Chem. 1999;274:32287–32294. © 1999 The American Society for Biochemistry and Molecular Biology.

In contrast, elevation of calcium by incubation of human lenses in A23187 caused a major increase in a spot migrating at a slightly lower molecular weight and more acidic position next to intact αB_1-175 in both the epithelium and the superficial cortex (Figs. 2A, 2B, A23187, arrows). This change was quite pronounced, in that few changes were observed in other crystallins. The novel protein was excised and digested and the peptides analyzed by tandem MS. These results indicate that the novel protein is derived by modification of intact αB_1-175. The modified αB was further analyzed by eluting it from a zinc-imidazole–stained 2-DE gel and measurement of its whole mass by electrospray mass spectrometry. After elution from a 2-DE gel, intact human αB had a measured mass of 20,201 (Fig. 3A), which was identical with the calculated mass. In contrast, the acidic and lower molecular weight form of αB produced by culture in A23187 had a mass that was 129 mass units lower (Fig. 3B), indicating that this protein was αB_1-174 missing a C-terminal lysine. The 2-DE migration and MS identification of truncated αB_1-174 was the same as previously described in some cataractous human lenses (Fig. 3C). Thus, the elevation of calcium in cultured human lenses reproduced a major protein change observed in vivo in human cataracts. 

Figure 3.

 

Deconvoluted spectra showing the masses of intact αB_1-175 (A) and truncated αB_1-174 (B) from human lens; and intact αB_1-175 (C) and truncated αB_1-174 (D) from monkey lens (spots with arrows in Figs. 2, 4).

Monkey Lens Culture in Calcium Ionophore

To test whether the formation of truncated αB_1-174 could be reproduced in nonhuman primate lenses, we conducted similar experiments in monkey lenses. Similar to human lens, culturing monkey lenses from animals 6-, 8-, and 12-years of age in A23187 caused a major increase in a spot migrating at a slightly lower molecular weight and more acidic position next to intact αB_1-175 (Fig. 4, arrows). Intact M. mulatta αB_1-175 was eluted from a 2-DE gel and its mass measured (Fig. 3C). Within experimental error, its 20,187 mass was identical with the calculated mass of M fascicularis αB_1-175 (Swiss Prot accession no. Q60HG8), suggesting that the sequences of intact αB are identical in these two species. The new spot migrating at a lower molecular weight and more acidic position compared to αB_1-175 had a mass of 20,059 (Fig. 3D). This 128-mass decrease was again due to the loss of the C-terminal Lys₁₇₅ from αB. Thus, monkey lenses provide a useful model for the study of the response of human lenses to calcium loading due to culture in A23187. 

Figure 4.

 

2-DE gels of the soluble proteins from the superficial region of monkey lens cortexes of ages 6, 8, and 12 years cultured without (Normal) or with A23187. Arrows: increased amounts of truncated increased αB_1-174 observed only in A23187-treated lenses.

CPE Activity against αB_1-175

Three carboxypeptidases (E/H, D, and X) are known to be expressed in human lens. ¹⁶ Commercially available human recombinant carboxypeptidase E (hrCPE) was used to digest human recombinant αB-crystallin (αB_1-175). Before the incubation study, CPE expression in monkey lens tissues was confirmed by immunoblot analysis (Fig. 5A). CPE was expressed only in the lens fiber region and not in the lens epithelium. This band did not appear if the CPE antibody was first neutralized with the immunizing peptide. These data indicated that the strong 50-kDa band in lens fiber protein was indeed CPE. αB_1-175 was truncated in a human lens epithelium sample (Fig. 2A), perhaps because of lens cortex contamination of the lens epithelium during dissection. The A23187-treated lenses absorbed significant amounts of water, which caused difficulties in dissecting the epithelium without fiber contamination. 

Figure 5.

 

Carboxypeptidase E immunoblot image of monkey lens epithelium (Epi) and fiber region (Fiber) lysate (A). Deconvoluted spectra showing the masses of recombinant αB_1-175 incubated with CPE and 2.5 μM Zinc (B), αB_1-175 incubated with CPE and 4 mM Calcium (C), αB_1-175 incubated with CPE, 2.5 μM Zinc and 4 mM Calcium (D), and αB_1-175 incubated with CPE and no ions (E).

CPE is known as a zinc metallocarboxypeptidase that removes basic amino acids from the C terminus of peptides. ³¹ We tested to determine whether Ca²⁺ would also activate CPE and hydrolyze αB-crystallin. γD- and γS-crystallin were used as negative control substrates since the C termini do not contain a basic amino acid. βB1-crystallin was used as another negative control because it has a Pro-Lys C terminus, and other basic carboxypeptidases did not hydrolyze substrates having a penultimate proline. ^32–35 When αB-crystallin was incubated for 1 hour with CPE in the presence of either calcium or zinc, an αB-crystallin fragment at 20,030 missing the C-terminal lysine was detected. Zinc (2.5 μM) caused 86% αB degradation (Fig. 5B), 4 mM calcium caused 33% degradation (Fig. 5C), and 2.5 μM zinc plus 4 mM calcium caused 93% degradation (Fig 5D). Zinc (4 mM) with CPE caused 100% degradation of αB (data not shown). All αB degradations were inhibited 100% by 1 mM 1,10-penanthroline chelating reagent (data not shown). 1,10-Penanthroline was a better inhibitor of CPE for chelating zinc ions than was EGTA. As we expected, none of the negative controls, βB1-, γD- and γS-crystallin, were degraded by CPE, even with 4 mM zinc (data not shown). These data confirmed that calcium, as well as zinc, is an activator of CPE. 

Unexpectedly, αB was degraded 19% by CPE without external addition of calcium or zinc, even after pretreatment with 2 mM EGTA (Fig. 5E). The assay buffer did not contain zinc; however, the CPE+αB mixture was found to contain 0.89 ± 0.02 mEq Zn/kg water (Table 3). This suggests that CPE binds zinc strongly enough to prevent removal during pretreatment with EGTA and to allow partial activation without addition of ions. 

Table 3.

 

Ion Concentrations and Water Levels during Monkey Lens Culture and Enzyme Assays

Sample	Zinc	Calcium*	Water
	(mEq/kg water)		(mg)
Lens (no culture)	0.48 ± 0.06	2.57 ± 0.14	101 ± 4
Lens (cultured)	0.42 ± 0.04	4.88 ± 1.69	113 ± 4
A23187 lens (cultured)	0.28 ± 0.03	9.91 ± 1.07	153 ± 8
CPE assay buffer	ND
CPE + αB	0.89 ± 0.02
EMEM medium	0.07 ± 0.02

 Data are means ± SEM (n = 3 for ion concentrations, n = 6 for lens water level). Three groups of nine lenses each were analyzed for ions and water.

*  Calcium data are referenced from a previous publication,(30) and are expressed as mEq/kg lens water, rather than as mmol Ca/kg lens water.

The A23187 ionophore used in the present lens culture experiments is known to increase influx of calcium and zinc from the medium. Only 0.07 ± 0.02 mEq Zn/kg water was detected in our culture medium (Table 3). Further, the zinc concentration actually decreased in A23187-treated lenses, probably due to the massive water uptake (Table 3). In contrast, A23187 increased lens calcium concentrations to almost two times over normal cultured lenses. Normal cultured lens showed a two-fold increase in calcium over initial lenses, possibly due to passive calcium movement during osmotic loading from the EMEM medium. A23187-treated lens showed even more calcium uptake than normal lens. These results suggest that endogenous CPE in the present experiments was activated by the more pronounced A23187-induced calcium influx. 

Discussion

Elevated amounts of truncated αB_1-174 have been observed in cataracts of some patients, ¹¹ in Soemmerring's ring after-cataract, ³⁶ and in an infant with persistent hyperplastic primary vitreous, in which cells invade the posterior lens. ¹¹ A major finding of the present investigation was that elevation of lenticular calcium by A23187 also produced αB_1-174 in cultured human lenses and that the process could be reproduced in cultured monkey lenses. However, calcium elevation did not reproduce the previously observed 16.4-kDa fragment of αB. ¹¹  

Our findings are important because the cause of elevated αB_1-174 has been unknown, and the data now implicate elevated calcium in the mechanism for production of αB_1-174. The findings in an earlier study ruled out premature termination of the αB-crystallin gene, splice variants, or other mutations as causes. ¹¹  

The comparative 2-DE maps of protein identities in human and monkey lenses produced in this study indicated that the positions of the various crystallins were remarkably similar. The major difference was that monkey lenses contained a greater number of acidic forms of all the major crystallins, especially αB-crystallin. Although a major portion of this acidification was probably due to the accumulation of deamidation in these long-lived proteins, ⁴ crystallins in monkey lenses also appeared to undergo a more extensive phosphorylation than human crystallins. The train of αB spots with increasing acidity in Figures 1C and 1D exhibited decreased migration in the second dimension compared with that of intact αB. This would be expected if they were phosphorylated. While 2-DE maps of human lens crystallins have been published, ^12,22,25 they were either constructed by using 2-DE gels focused with free carrier ampholines, or were IPG based with less complete spot identification. The comparative 2-DE maps between human and monkey lenses in this study should facilitate the use of monkey lenses to investigate the contribution of crystallin modifications in the formation of human cataract. 

The further usefulness of cultured monkey lenses to study human lens biology was demonstrated by the similar manner in which both species lost the C-terminal Lys of αB-crystallin. We hypothesize that this cleavage was probably due to carboxypeptidase activation by calcium, rather than zinc, influx into the lens. However, note that our data showed a much greater sensitivity of CPE to the zinc ion, even with pretreatment with EGTA. Normal lens contains a significant level of zinc (∼1 mM), and zinc was not elevated in our A23187-treated lenses. Zinc binds to approximately 3000 proteins, especially 200 enzymes and 200 transcription factors. ³⁷ Thus, zinc may have been bound to these other proteins in normal lens, leaving too little to activate carboxypeptidases. In calcium-elevated lenses, however, calcium may directly bind to carboxypeptidases for activation. The specific cleavage of αB may result from the nearly complete absence of basic residues at the C-terminal end of other crystallins. Although βB1-crystallin also contains a C-terminal Lys residue, an adjacent Pro residue prevented cleavage. Similarly, Lys 174 of αB-crystallin may be resistant to further cleavage due to Pro at residue 173. 

The functional significance of truncated αB_1-174 in lens remains unknown. Although intact αB-crystallin plays an important role as a chaperone in maintaining the solubility of other proteins in aged lens, ⁶ removal of the C-terminal lysine does not appear to alter chaperone activity. ¹¹ However, it is possible that other functions, such as those leading to genomic instability and hyperproliferation in αB^−/− mice ¹⁴ is affected by this truncation. Alternatively, the αB truncation may have no functional significance, but could be an indicator that unregulated carboxypeptidase activity in calcium loaded or stressed lens cells alters the function of other proteins that then could contribute to cataract. For example, of the 20,332 protein entries listed in the Swiss-Prot database (ver. 57.2), 1,425 contain a C-terminal lysine without a penultimate proline. Some of these are also likely targets for unregulated carboxypeptidase activity. 

Human lens is known to contain calpains, ³⁸ a class of proteases activated by calcium. In an earlier study, these enzymes were implicated in the proteolysis of the cytoskeletal protein vimentin in human lenses cultured with ionomycin. ¹⁷ However, the present experiments and our earlier study ³⁰ suggest that the activity of calpains in human and monkey lenses are not significantly directed toward the crystallins. Calpains are known to extensively truncate the C terminus of both αA- and αB-crystallin. ³⁹ The lack of any observable truncation of αA- or the more extensive truncation of αB-crystallin suggests that elevation of calcium by A23187 in human and monkey lenses does not lead to widespread activation of calpains. This apparent lack of calpain activity against the crystallins in human and monkey lenses is due to the low activity of the calpains, and the relatively high abundance of calpastatin, the endogenous inhibitor of calpains, in these lenses. ^30,38 Also the absence of Lp82, the isoform of calpain 3, in primates is a major difference. Lp82 is also much more resistant to calpastatin, and Lp82 degrades calpastatin. ⁴⁰ The lack of calpain activation was unlike that in rodent lenses, which undergo extensive opacification and calpain-induced truncation of crystallins during culture in A23187. ⁴¹ The results in the present study also suggest that calcium elevation by A23187 does not cause extensive transglutaminase mediated cross-linking of crystallins, since no cross-linking of βA3-crystallin, a known substrate for lens transglutaminase, ⁴² was observed. These results were unlike those in a previous study of human lenses cultured in ionomycin that observed possible transglutaminase cross-linking of vimentin. ¹⁷ Thus, unlike vimentin, crystallins in primate lenses appear to be resistant to modification by both calpains and transglutaminase. 

Further studies are necessary to investigate why calcium elevation leads to carboxypeptidase activity. These studies are important, because inappropriate carboxypeptidase activity could contribute to the mechanism of human cataract and potentially be a target for therapeutic intervention. 

Supplementary Materials

Supplementary Data S1 - (PDF) 

Supplementary Data S2 - (PDF) 

Footnotes

 Supported by NIH Grants EY07755 (LLD), EY05786 (TRS), and Core Grant EY10572. TRS is a paid consultant for Senju Pharmaceutical Co., Ltd., a company that may have a commercial interest in the results of this research and technology. MA is and EN was (during the study) an employee of Senju Pharmaceutical Co., Ltd. This potential conflict of interest was reviewed, and a management plan approved by the OHSU Conflict of Interest in Research Committee was implemented.

Footnotes

 Disclosure: E. Nakajima, Senju Pharmaceutical Co., Ltd. (E, F); L.L. David, Senju Pharmaceutical Co., Ltd. (F); M.A. Riviere, Senju Pharmaceutical Co., Ltd. (F); M. Azuma, Senju Pharmaceutical Co., Ltd. (F); T.R. Shearer, Senju Pharmaceutical Co., Ltd. (C, F)

Footnotes

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

The authors thank Phillip Wilmarth for technical assistance and Ninan J. Blackburn (Department of Environmental and Biomolecular Systems, Oregon Health and Science University) for use of the ICP instrument for zinc determination.