Characterization of a Bradykinin-Hydrolyzing Protease from the Bovine Lens

Abstract

purpose. To isolate and characterize bovine lens endopeptidase activity that cleaves the Phe-Ser bond in peptide substrates.

methods. The protease activity in young bovine lens homogenate was measured using the Mca-(Ala⁷,Lys(Dnp)⁹)-bradykinin substrate. Degradation of bradykinin and other unlabeled peptide substrates was monitored by reversed-phase HPLC on a C18 column. The protease was purified by means of several chromatography steps. An in-gel tryptic digest of the purified protease was analyzed by using matrix-assisted desorption ionization-time of flight mass spectrometry (MALDI-ToF-MS and nanospray quadrupole time of flight mass spectrometry (QqToF MS). The specificity of the protease was determined with bradykinin and its analogues. Crystallin fragments isolated from aged bovine lenses were tested for their susceptibility to degradation by a newly identified endopeptidase.

results. Bradykinin hydrolyzing endopeptidase activity was localized mainly in the outer cortex of the lens. A characterization study showed that the purified protease was thiol and metal dependent. Peptide mass fingerprinting and tandem mass spectrometry (MS/MS) analysis of an in-gel tryptic digest matched the protein sequence of thimet oligopeptidase (TOP). The purified protease cleaved bradykinin specifically at Phe⁵-Ser⁶ and neurotensin at Arg⁸-Arg⁹. Basic or hydrophobic amino acids at P1 and P1′ positions in the substrate were preferred over acidic residues. Enzyme activity was also inhibited by physiological levels of adenosine triphosphate (1 mM; ATP) and glutathione (3 mM; GSH). The crystallin fragments obtained from aged bovine lenses were cleaved by the purified enzyme.

conclusions. This study shows the presence of TOP in the bovine lens. Its unique substrate specificity and regulation of its activity by ATP and GSH suggest that TOP has an important role in peptide hydrolysis in the lens.

Degradation of lens crystallins by proteolytic enzymes can damage their highly ordered arrangement in lens, leading to loss of lens transparency. Compared with normal lenses, cataract lenses exhibit more fragmentation of lens proteins due to the action of proteolytic enzymes. ^{1 2 3} It has been hypothesized that the resultant crystallin fragments (short polypeptides) interact with the lens proteins in vivo, leading to the increased formation of high-molecular-weight aggregates and scattering of light. Several proteolytic enzymes have been shown to be associated with aging and cataractous changes in the lens. ^{1 4 5} Enzymatic cleavage of proteins is essential for lens fiber cell differentiation and organization. ⁶ Whereas truncation of β-crystallin, vimentin, spectrin, and membrane proteins has been attributed to the initial tight packing of crystallins during lens maturation, ⁶ proteolysis has been implicated in aging and cataract. ² Activation of calpains during the process of cataractogenesis has been extensively studied in human ⁷ and mouse models. ³ N-terminal extension of β-crystallins is cleaved by rodent calpains. ² Truncation of α-crystallin by proteases reduces its chaperone-like property. ⁸ Therefore, it can be argued that proteases play a critical role in cataractogenesis through their action on crystallins. 

Multicatalytic protease plays an important “housekeeping” role in the lens epithelial layer, and its activity is lower in human cataract lenses than in normal lenses. ⁹ In earlier studies, we purified a prolyl oligopeptidase, ¹⁰ an acylpeptide hydrolase, ¹¹ and a puromycin-sensitive amino peptidase ^{12 13} from bovine lenses. The lens is also a rich source of leucine aminopeptidase. ¹¹ Recently, a bestatin-sensitive aminopeptidase was partially purified from rat lens cortex and shown to be involved in fiber cell globulization and lens opacification. ¹⁴ Endoproteases such as calpains and proteasomes are involved in the proteolysis of crystallins in the lens, and crystallin fragments are further attacked by exopeptidases and endopeptidases. Because some of the modified crystallin fragments have the ability to interact with other crystallins and form components of water-insoluble aggregate, ¹⁵ proteases and peptidases that can degrade crystallin fragments into smaller peptides and amino acids would provide an important secondary defense in vivo. Recently, we have shown that crystallin fragments obtained from oxidized β-l-crystallin significantly increase thermally induced aggregation of lens proteins. ¹⁵  

Based on the degradation of peptides in the in vivo cleavage sites of α-crystallin, we hypothesize that novel proteases in the lens recognize a specific peptide sequence around cleavage sites and may be responsible for crystallin cleavage during aging and cataractogenesis. ¹⁶ Bovine lens extract was shown to degrade the N-terminally blocked peptides containing the peptide bonds Thr-Ser, Ser-Ala, and Ser-Ser, which represents in vivo cleavage sites in α-crystallin. This newly detected protease activity was resistant to diisopropyl fluorophosphate (DFP) and E64 but sensitive to thiol-blocking agents. ¹⁶ The present study describes the identification and characterization of an endopeptidase (thimet oligopeptidase; TOP) from young bovine lenses that specifically degrades Phe-Ser bond in peptide substrates. 

Materials and Methods

Tissue

Fresh young and old bovine lenses were purchased from Pel Freez Biologicals (Rogers, AR), and human lenses were obtained from the Lion’s Eye Tissue Bank of Missouri (Columbia, MO). Lenses were stored at −70°C until use. The thawed lenses were decapsulated and homogenized by stirring in 50 mM Tris-HCl (pH 7.0), containing 1 mM dithiothreitol (DTT) at 4°C. 

Detection of Phe-Ser–Cleaving Activity in Lens Extract

Bradykinin contains several cleavage sites for identifying different proteases. ¹⁷ To assess bradykinin degradation, 50 μL of dialyzed lens extract (∼4 mg protein) was incubated with 50 μg bradykinin (Sigma-Aldrich, St. Louis, MO) in 200 μL of 10 mM Tris buffer (pH 7.5), 100 mM NaCl and 0.5 mM DTT. Bestatin (1 mM, Sigma-Aldrich) was included in all assays to inhibit the aminopeptidases present in the extracts. After 3 hours of incubation, the diluted incubation mixture was filtered through 10-kDa cutoff centrifuge filters (Amicon; Millipore, Bedford, MA). The peptides present in the filtrates were analyzed by HPLC using a C18 reverse-phase analytical column (Grace Vydac, Hesperia, CA). To separate the peptides, a linear gradient of acetonitrile, containing 0.05% trifluoroacetic acid (Sigma-Aldrich) from 0% to 10% for 15 minutes followed by 10% to 40%, was run over 40 minutes at a flow rate of 1 mL/min. In a separate experiment, lens extract was preincubated (30 minutes) with 0.5 mM DFP (Sigma-Aldrich) and 0.5 mM l-trans-epoxysuccinyl-l-leucylamido-4-(guanidino) butane (E64; Sigma-Aldrich). Incubation was continued for 3 hours after the addition of bradykinin, after which the reaction products were separated. The peptide peaks detected at 220 nm were collected separately and analyzed by mass spectrometry. 

Internally Quenched Fluorogenic Substrates

Protease activity in bovine and human lens homogenate was also measured by using an internally quenched bradykinin substrate, Mca-(Ala⁷,Lys(Dnp)⁹)-bradykinin (IQS-BK; Bachem, Torrance, CA). Substrate (3 nanomoles) was added to 200 μL of 20 mM Tris/HCl (pH 7.4), 100 mM NaCl, 0.5 mM β-mercaptoethanol (BME), and 0.05% Brij35 and the activity measurements at 25°C were initiated by the addition of 50 μL of purified protease or protease fraction obtained during different purification steps. Fluorescence was measured in a 96-well plate with a plate reader (FLx 800; Bio-Tek, Winooski, VT), with excitation at 340 ± 30 nm and emission at 395 ± 25 nm. 

Purification of Protease

Young bovine lenses were decapsulated and homogenized at 4°C in 10 mM sodium acetate-acetic acid (5 mL/lens; pH 5.2), at 4°C. ¹⁸ The homogenate obtained from 60 lenses was readjusted to pH 5.2 with 1 M acetic acid and then centrifuged for 30 minutes at 10,000g. The supernatant was dialyzed against 10 mM Tris-HCl buffer (pH 7.5), 0.5 M NaCl, and 5 mM imidazole (buffer A) for 12 hours. 

Protease activity was monitored during purification steps, using IQS-BK as the substrate. The dialyzed sample was passed through resin (His-Bind; Novagen, Madison, WI), to eliminate the postproline endopeptidase and aminopeptidase activity present in the lens extracts. ¹⁹ Wash fractions that contained protease activity were concentrated in an ultrafiltration system (Amicon, Millipore) and dialyzed against 10 mM Tris-HCl buffer (pH 7.5) containing 0.05% Brij35 and 2 mM BME (buffer B). The preparation was then run on an anion exchange column (Macroprep High Q column; Bio-Rad, Hercules, CA; 60 mL bed volume), which had been equilibrated with buffer B. Elution was performed with a linear gradient of salt, up to 0.5 M NaCl, in buffer B over five bed volumes. The peak fractions sensitive to Z-PheΨ (PO₂CH₂)-Ala-Arg-Phe, a phosphinic peptide inhibitor of TOP (a gift of Vincent Dive; CEA, Saclay, France) were pooled and exchanged by dialysis against 10 mM sodium phosphate buffer (pH 7.0). The protease fraction was then passed through a column of hydroxyapatite (40 mL, Bio-Gel HTP; Bio-Rad) in 10 mM sodium phosphate buffer (pH 7.0) with a gradient of up to 200 mM sodium phosphate (pH 7.0) over six bed volumes. The active fractions were pooled and dialyzed against 20 mM phosphate (pH 7.5) and 0.5 M NaCl and run on a second resin column (His-Bind; Novagen) without imidazole. The wash fractions were concentrated by ultrafiltration, exchanged to buffer B and then run on an ion-exchange column (Protein Sax 300 VHP575P; Grace Vydac) with an NaCl gradient up to 125 mM. To obtain the final product, the sample was concentrated to 0.5 mL and passed through preparation-grade gel (100-mL bed volume; Superdex-200; Amersham Biosciences, Piscataway, NJ). 

Identification of TOP

The unique substrate specificity of TOP was studied by using bradykinin, neurotensin, and adipokinetic hormone G peptide substrates. The TOP-specific cleavage sites in bradykinin and neurotensin (Sigma-Aldrich) were confirmed by mass spectrometric analysis of degradation products purified using HPLC-C18 columns. The identity of the protease was further confirmed by using specific phosphinic peptide inhibitor Z-PheΨ (PO₂CH₂)-Ala-Arg-Phe. The purified protease (5 μg) was preincubated for 30 minutes with 1 μM phosphinic peptide inhibitor, and then the protease activity was measured. 

Silver-stained bands (GelCode; Pierce, Rockford, IL) were subjected to in-gel trypsin digestion ²⁰ and analyzed by matrix-assisted desorption ionization-time of flight mass spectrometry (MALDI-ToF; Voyager-DE Pro; Applied Biosystems, Foster City, CA), and the spectra were analyzed by computer (MS-Fit program of Protein Prospector; available at http://prospector.ucsf.edu, hosted by the University of California at San Francisco ²¹ ). The selected peptides were further analyzed by nanospray quadrupole time of flight mass spectrometry (QqTOF MS; on a Sciex QStar/Pulsar i instrument [Applied Biosystems/MDS] fitted with a Protana [Odense, Denmark] nanospray source). 

Degradation of Lens Peptides

Peptides from aged bovine lenses were separated by sonicating the lens extract in the presence of 4 M guanidine hydrochloride and removing the soluble proteins by 5% trichloroacetic acid precipitation. The supernatant containing the peptides was filtered through a 10-kDa filter, and the filtrate was dialyzed extensively against 10 mM phosphate buffer (pH 7.5) using a 0.5-kDa membrane. Then, 200 μL of peptide solution (obtained from five old lenses) containing nearly 1 nanomole of peptide was incubated with 2 μg of purified protease for 1 hour at 37°C. The amount of free amino groups released was measured by the O-phthalaldehyde method. ²² Purified TOP was also incubated separately. The difference in the amino groups before and after incubation constituted a measure of degradation of lens peptides. 

Characterization of the Protease

Protease activity was estimated in the presence of serine, cysteine, aspartate, and metalloprotease inactivators as well as aminopeptidase inhibitors. Purified protease was preincubated with different inhibitors for 30 minutes under the experimental conditions described previously. We designed an irreversible inhibitor, botinylated-BK(1-5)-chloromethylketone, to assess the specificity of TOP. Both crude lens extract and purified TOP were preincubated with 1 mM DTT, followed by incubation with 50 μM biotinylated-BK(1-5)-chloromethylketone (American Peptide Company, Sunnyvale, CA). The residual protease activity was measured with bradykinin or IQS-BK used as substrate. 

To determine the extent of inhibition of TOP activity by ATP, aliquots of crude lens extract and purified protease were preincubated with 0 to 10 mM ATP in 10 mM Tris-buffer (pH 6.9). Different concentrations of DTT and GSH in Tris-HCl buffer (pH 6.9) were preincubated with the purified protease, and degradation of IQS-BK was measured. For optimum pH studies, Tris-imidazole buffers (20 mM; pH 5.3–8) were used. 

Results

Degradation of Bradykinin in Bovine Lens Extract

To assess bradykinin degrading endopeptidases in bovine lens extract, water-soluble bovine lens extract was incubated with bradykinin in the presence of bestatin (0.1 mM). Bradykinin was cleaved at different sites by proteases present in the lens extract (Fig. 1A) . The products from bradykinin hydrolysis were identified by mass spectrometry. They were formed by the cleavages on the carboxyl side of Pro², Phe⁵, Ser⁶, and Pro⁷ of bradykinin. Bovine lens contains a postproline cleaving protease, prolyloligopeptidase (POP), which can readily cleave peptides on the carboxyl side of Pro. ¹⁰ Therefore, the crude lens extract was treated with 0.1 mM DFP to inactivate the POP, and the ability of the residual protease activity to hydrolyze bradykinin was determined. Only two peptides were generated from bradykinin due to the hydrolysis of Phe⁵-Ser⁶ bond in bradykinin (Fig. 1B) . The peaks c and e were found to be peptides SPFR (m/z, 507.05) and RPPGF (m/z, 574.21), respectively. In contrast, control lens extract by itself showed no release of these peptides during incubation. 

Purification of the Endoprotease from Young Bovine Lens

Endopeptidase activity occurred mainly in the outer cortex of the young bovine lens (Fig. 2) . The innermost cortical fibers showed low TOP activity, but the nuclear region was almost devoid of TOP activity. Endopeptidase was purified to homogeneity by using different chromatography steps. The major portion of the crystallins was separated from TOP using an immobilized metal ion affinity chromatography (IMAC) column. ¹⁹ The concentrated sample containing TOP activity was then dialyzed extensively against the low ionic-strength buffer and fractionated on an anion exchange column (Macroprep High Q column; Bio-Rad) using 0 to 500 mM salt gradient. All fractions were tested for IQS-BK degradation, and the fraction containing protease activity was tested for its substrate specificity by using intact bradykinin. At this stage, the Phe-Ser–cleaving protease fraction was free from other bradykinin-degrading enzymes. Protease was eluted at two gradient ionic concentrations, between 100 and 125 mM NaCl (Fig. 3A) . Both activity peaks 1 and 2 degraded bradykinin at Phe⁵-Ser⁶, but peak 1 showed relatively less activity, and it was not inhibited by the TOP-specific phosphinic peptide inhibitor. Peak 2 was maximally inhibited (up to 70%) by 2 μM phosphinic peptide inhibitor. Protease activity in peak 2 was further purified on a second nickel column without imidazole. Phosphinic peptide inhibitor at this stage showed more than 90% inhibition of activity. Hence, the resultant protease was further purified on an HPLC column (Fig. 3B ; 300VHP575; Grace Vydac). Final purification was achieved with a second column (Superdex G-200; Amersham Biosciences). In the end, 25,000-fold purification was attained with specific activity of 0.75 U/mg. The apparent molecular mass was ∼79 kDa on both gel filtration and SDS-PAGE (Fig. 4) . The results of the purification procedure are summarized in Table 1 . For bradykinin substrate, the K _m was 6.25 μM. The pH profile showed that lens TOP is optimally active at neutral pH. 

Mass Spectrometric Identification of TOP

The purified protease was detected on 10% SDS-PAGE gel using silver stain (GelCode; Pierce). The peptide mixture generated by in-gel trypsin digestion was analyzed with MALDI-ToF MS. ²³ Automated analysis of peptide mass fingerprinting was performed in an autofit program on computer (Protein Prospector Auto MS-Fit; PE Biosystems; Fig. 5A ). The tryptic peptides of lens protease matched 35% of the porcine TOP (P47788) sequence (Fig. 5B) . Seven major peptide peaks were analyzed on nanospray QqToF. MS/MS of all seven abundant peptides from the digest showed that lens protease is homologous with porcine liver TOP (Table 2) . MS/MS spectra acquired for the individual peptide were analyzed using the probability-based Mowse score (P < 0.05) in Mascot software (Matrix Science, London, UK). Figure 6 shows the nanospray MS/MS of the peptide ⁶⁷ALADVEVSYTVQR⁷⁹ with M+H⁺ 1450.83. 

Effect of Inhibitors

In this study, lens TOP activity was completely inhibited by p-chloromercuribenzoate (PCMB), 1,10-phenanthroline, and biotinylated bradykinin(1-5)-chloromethyl ketone (Table 3) . A potent phosphinic peptide inhibitor of TOP, Z-PheΨ (PO₂CH₂)-Ala-Arg-Phe inhibited TOP activity completely. TOP activity was not inhibited by any of the general serine and cysteine protease inhibitors, such as DFP, E64, N-p-tosyl-l-phenylalanine chloromethyl ketone (TPCK), Nα-p-tosyl-l-lysine chloromethyl ketone (TLCK), and lactacystin. N-ethylmaleinide (NEM), EDTA, and iodoacetamide exerted partial inhibition. EDTA exerted less inhibition at 1 to 2 mM concentration than O-phenanthroline, indicating that metal ion is tightly bound to the protease. Incubation with 1,10-phenanthroline (1 mM) led to complete loss of protease activity. Metal ions such as Zn²⁺ and Ca²⁺ induced concentration-dependent inhibition of TOP activity, with complete loss of activity between a 1- to 2-mM concentration (Table 3) . Because of the same substrate specificity, neurolysin activity could be mistaken for TOP activity. ²⁴ Therefore, we tested the enzyme activity in the presence of Pro-Ile, a specific inhibitor of neurolysin. Pro-Ile (5 mM) had no effect on the lens TOP activity, suggesting that the preparation was free of neurolysin. Brain TOP has been shown to have variable degrees of susceptibility to thiol-blocking agents. ²⁵ In our study, glutathione and DTT showed similar effects on IQS-BK and bradykinin degradation by the purified protease. Whereas a low concentration of DTT (0.5 mM) increased TOP by 1.5-fold, and a higher level (5 mM) produced a more than 60% inhibition of bradykinin degradation. Similarly, the lens protease was maximally activated by 1 mM GSH, a higher GSH concentration decreased the activity (Fig. 7) . Therefore, bovine lens TOP properties are characteristic of TOP from other tissues. ²⁵  

Substrate Specificity of Purified Protease

The site of cleavage in bradykinin, neurotensin, adipokinetic hormone G, and IQS-BK substrates was analyzed by HPLC followed by MALDI-ToF MS (Table 4) . The peptide bond between the Phe and Ser residues was preferentially cleaved in the case of bradykinin and adipokinetic hormone G, but in the case of neurotensin, the peptide bond between Arg-Arg was preferentially cleaved. The substrate specificity of the purified protease was also examined using synthetic d-alanyl-bradykinin analogues with substitutions at the P1 to P1′ positions (Table 5) . P1 and P1′ represent the amino acids contributing the amino group and carboxyl groups of the susceptible peptide bond in a substrate. The enzyme showed maximum activity with (d)ARPPGYSPFR and (d)ARPPGLSPFR substrates. When the P1 position of BK was replaced by hydrophobic or basic amino acids, an increase in TOP activity was observed. No hydrolysis was seen when P1 or P1′ residues were replaced by acidic amino acids. 

Effect of ATP

An earlier study showed that Zn²⁺ at the active center of TOP is the target for ATP binding, resulting in competition for substrate binding. ²⁶ TOP also undergoes autophosphorylation. Compared with the other tissues, lens tissue contains a relatively higher amount of ATP. ²⁷ It can therefore be speculated that TOP activity in lens is regulated by ATP. Our study showed significant inhibition of TOP activity by a physiological concentration of ATP (Fig. 8A) . Protease activity in crude lens extract was inhibited by ATP in a concentration-dependent manner. Figure 8B shows the inhibition of purified protease by 1.0 mM ATP. 

Degradation of Lens Peptides

A previous study showed that peptides tend to accumulate with proteins in the nuclear regions, ²⁸ whereas the proteases show higher activity in the outer cortical region. When the peptides isolated from the inner cortex and nuclear region of old bovine lenses were subjected to hydrolysis by purified TOP in the absence of phosphinic peptide inhibitor a significant increase occurred in the new amino groups, indicating that TOP is capable of degrading such peptides from lens (Fig. 9) . It is likely that the diminished activity of TOP in aged lenses (and in the lens nuclear region) may be responsible for the accumulation of peptides in vivo. 

TOP Activity in Bovine versus Human Lenses

TOP activity was significantly decreased in aged bovine lens extract compared with that of young bovine lens extracts (P < 0.001). Aged (60–80 years) human lens extracts showed only a marginal decrease in TOP activity compared with relatively younger (18–25 years) human lens extracts (Table 6) . Long-term storage of lenses, however, did not seem to be a factor contributing to this difference. Because all the human lenses (young and old) analyzed were in fact older than the aged bovine lenses, it is plausible that a larger difference between young and old human lenses was not observed. 

Discussion

In our study of bovine lens, an endopeptidase that specifically cleaves bradykinin at Phe⁵-Ser⁶ was isolated and characterized. Because our initial observations showed specific cleavage of bradykinin, it was chosen as the substrate during a lens TOP isolation and characterization study. This metal- and thiol-dependent protease activity was predominantly found in the outer cortex of bovine lens. Bradykinin is rapidly degraded by lens prolyl oligopeptidase and aminopeptidases, but by using bestatin and DFP, we were able to detect the new protease activity within 1 hour of incubation of crude lens extract with bradykinin and IQS-BK. Much of the copurifying β-crystallin and prolyl endopeptidase was separated from the lens extract using an IMAC column. Peptide mass fingerprinting and MS/MS analysis of the purified protease showed significant sequence homology to porcine TOP (Swiss Prot P47788, ∼79 kDa; http//:www.expasy.org/; hosted by the Swiss Institute of Bioinformatics, Geneva, Switzerland). 

The complete inhibition of enzyme activity by 1,10-phenanthrolene and characteristic inhibition by zinc and calcium observed during this study is in agreement with previous studies on TOP. ^{25 29} Zinc-dependent metalloproteases play a vital role in the complete degradation of modified and abnormal proteins. Several studies on substrate specificity have shown that TOP preferentially cleaves the peptides that have 6 to 17 amino acids. ^{18 30} In this study, we show that lens protease cleaves the peptide-containing hydrophobic amino acid or basic amino acid at the P1 position preferentially in the order Tyr>Leu>Phe>Arg. The rate of hydrolysis of substrate was lower whenever acidic amino acids contributed to the peptide bond (Table 5) . The specific phosphinic peptide inhibitor Z-PheΨ (PO₂CH₂)-Ala-Arg-Phe produced more than 90% inhibition of the TOP activity at less than 0.5 μM concentration (Table 3) . TOP follows stringent peptide selection for degradation, ³¹ indicating a specific role for this enzyme in the lens proteolytic system. Lew et al. ³² showed the evidence for a two-step mechanism for GnRH metabolism in the hypothalamus. They proposed that GnRH catabolism requires an initial cleavage by prolyl endopeptidase before the action by TOP. On a similar note, because lens is also a rich source of prolyl endopeptidase and aminopeptidases, net peptide metabolism of degraded polypeptides in lens may occur as a result of the combined action of endopeptidases and aminopeptidase in lens. 

Several studies have shown that accelerated proteolysis causes aggregation of crystallin fragments, leading to lens opacity. ^{4 5} The presence of increased amounts of crystallin fragments in water-insoluble proteins of bovine lens nucleus supports this view. ¹¹ Many peptidases, such as POP, LAP, AP-III, and ACP, have been reported in bovine lens, with all of them seeming to be limited or absent in the nuclear region. Several other studies have also indicated an active proteolytic mechanism in the truncation of lens crystallins. ^{33 34 35} In aging and cataract, lens proteases such as DPPII, DPPIII, and AAP-S are also expressed at high level in the perinuclear region. ¹ Further studies on the status of TOP in cataract lenses may provide additional information regarding the proteolytic cascade in cataractogenesis. 

A study of net proteolysis in glucose-deprived, intact, cultured rat lenses showed depletion of ATP and initiation of proteolysis. ²⁷ Earlier studies revealed that phosphorylation of TOP displays a dramatic increase in the value of K _m and K _cat for peptide substrates. ³⁶ We have also found that ATP (metal free) inhibits TOP activity both in crude lens extract and purified fraction. Earlier results suggest that Zn²⁺ at the active center of TOP is the target for ATP binding, which leads to inhibition of enzyme activity. ²⁶ DTT has been shown to have a dual effect: low levels (0.5 mM) activate TOP and higher levels (5 mM) inhibit the enzyme. ²⁶ Lens TOP also showed a similar profile with DTT and GSH. Inhibition has been attributed to the disruption of disulfide bridges within the enzyme and to the thiophilicity of the catalytic zinc ion. ³⁷  

Of note, a similar type of thiol-dependent zinc protease has been reported in the bovine lens capsule. ³⁸ The TOP-like sequence was shown to be buried in a basement membrane protein, fibronectin, which resolves into low-molecular-mass fragments (47 and 37 kDa) on SDS-PAGE. ³⁸ This protease is considered to be associated with such biological phenomena as maintenance of normal cell morphology, cell migration, and wound healing. TOP functions as an oligopeptide-hydrolyzing and -binding protein. TOP has chaperone-like activity, owing to its selective binding to ovalbumin oligopeptide without degradation. ³⁹ TOP has been shown to play an important role in processing of peptides formed from the proteasome activity. ^{26 40}  

Cataractous lenses have been found to contain less puromycin-sensitive aminopeptidase than normal lenses. ¹² In this study, TOP activity was also found to be significantly decreased in aged bovine lenses. Therefore, aged and cataractous human lenses can be expected to have more peptide accumulation than do young human lenses. The results of this study show, for the first time, that bovine lens contains TOP activity capable of cleaving a peptide bond contributed by hydrophobic amino acids in oligopeptides, thereby facilitating the complete degradation of hydrophobic peptides generated in vivo. The present data also show that lens TOP may be subject to regulation by ATP and an oxidation-reduction state in the lens. Further studies are under way to determine the in vivo role of TOP in the cleavage of crystallin fragments. 

 Supported in part by National Eye Institute Grants EY09855 and EY14795, the Lew Wasserman Merit award (KKS) and an unrestricted grant to the Department of Ophthalmology from Research to Prevent Blindness, Inc.

 Submitted for publication July 21, 2003; revised December 11, 2003; accepted December 19, 2003.

 Disclosure: R. Chaerkady, None; K.K. Sharma, None

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

 

Figure 1.

 

Degradation of bradykinin by bovine lens extract. Reversed-phase HPLC on a C18 column showed the separation of bradykinin fragments. (A) Lens extract (4 mg protein) was incubated for 30 minutes at 37°C with 50 μg of bradykinin in 10 mM Tris buffer (pH 7.5) containing 0.2 mM bestatin. (B) Same as (A) in the presence of 100 μM DFP and 0.2 mM bestatin. The peptide peaks were identified by MALDI-ToF MS. Peak a is FR; b is PFR; c is SPFR; d is PGF; e is RPPGF; f is RPPGFSP; g is bradykinin; and h is bestatin.

Figure 2.

 

Distribution of Phe-Ser hydrolyzing activity in bovine lens. Enzyme activity was measured using IQS-BK as the substrate in fractions obtained after gentle vortexing of the bovine lens in 50 mM Tris-HCl buffer (pH 7.0). The activity is expressed per total protein in each fraction (▪). Accumulated protein is shown on the second y-axis (□) and represents both water-soluble and water-insoluble lens proteins.

Figure 3.

 

Chromatographic purification of thimet oligopeptidase. (A) Fractionation of partially purified TOP activity on an anion exchange column (Macroprep High Q; Bio-Rad). The unbound protein fraction from resin (His-Bind; Novagen) was applied to a 2 × 30-cm anion exchange column, washed, and eluted with a salt gradient. Fractions (5 mL) were collected and assayed for protein at λ 280 nm (dashed line) and for TOP activity, using IQS-BK at pH 6.9 (solid line). Both peaks 1 and 2 hydrolyzed bradykinin at Phe-Ser bond, but only peak 2 was sensitive to specific phosphinic peptide inhibitor. (B) Fractionation of purified TOP by HPLC (Protein Sax 300 VHP575P column; Grace Vydec). The fraction from the second resin containing TOP activity was dialyzed against 10 mM Tris-HCl (pH 8.0) and passed through the HPLC ion exchange column with a gradient of NaCl from 0 to 250 mM. Absorbance was monitored at 280 nm. Fractions of 0.5 mL were collected and assayed by using IQS-BK.

Figure 4.

 

SDS-PAGE of lens TOP. The protease obtained from the separation column (Superdex 200; Amersham Biosciences) was reduced, denatured, and analyzed on 12% SDS-PAGE. Silver stain was used to localize the protein band. Open arrow: location of TOP in the stained gel.

Table 1.

 

Purification of TOP from Bovine Lens Cortex

Steps	Protein (mg)	Total Activity	Specific Activity*	Yield	Magnitude of Purification (×)
1. Lens extract water-soluble fraction	16,800.00	0.500	0.000030	100	1.00
2. After His-Bind resin affinity (imidazole, 5 mM; Novagen)	4,541.00	0.380	0.000085	76	2.80
3. After Macroprep High Q (peak 2) (Bio-Rad)	68.00	0.280	0.004150	55	138.00
4. Second His-Bind resin column (no imidazole; Novagen)	6.00	0.240	0.039200	46	1,306.00
5. Bio-gel HTP (hydroxyapatite column; Bio-Rad)	0.75	0.180	0.239000	35	7,966.00
6. After 300 VHP575P HPLC column (Grace Vydac)	0.28	0.168	0.600000	33	20,000.00
7. After Superdex 200 (Amersham Biosciences)	0.04	0.030	0.750000	5	25,000.00

*  The isolation procedure was started with 60 bovine lenses. The activity (units/milligram) was determined using Mca-(Ala⁷,Lys(Dnp)⁹)-bradykinin (IQS-BK) at 25°C and pH 6.9.

Figure 5.

 

Peptide mass fingerprinting of TOP. The 79-kDa protein band on 12% SDS-PAGE was digested using sequence-grade trypsin and analyzed in MALDI-ToF MS, as explained in the text. (A) MALDI-ToF MS of the tryptic digest peptides. Inset: observed mass of the peptides. (B) The region of the peptide sequence was obtained by peptide mass mapping. The calculated mass (monoisotopic mass, M+H⁺) is shown under the corresponding peptide. Peptides with cysteine residues were considered to be modified with acrylamide adducts (+71 Da).

Table 2.

 

Tryptic Peptides of TOP Analyzed by Nanospray Mass Spectrometry

m/z Observed (Da)	z	Calculated M+H+ (Da)	Fragment of TOP Matched
453.21	2	905.41	INAWDMR (332–338)
458.27	2	915.53	IVWLQEK (122–128)
567.31	2	1133.61	QANTGLFNLR (544–553)
635.32	2	1269.63	EYFPMQVVTR (358–367)
725.88	2	1450.75	ALADVEVSYTVQR (67–79)
895.44	2	1789.87	EELGGLPEDFLNSLEK (194–209)
887.10	3	2659.28	DFVEAPSQMLENWVWEAEPLLR (499–520)

 Seven peptides from the MALDI-ToF spectrum were further analyzed by nanospray QqToF MS/MS. The doubly and triply charged parent ions were selected for the production of CID spectra. The calculated mass of the peptide is the monisotopic (M+H⁺) mass for those peptides. Residue number is shown in parentheses.

Figure 6.

 

MS/MS spectrum of tryptic peptides 67-79 of bovine lens TOP. The measured masses of the identified y ions (loss of residues at the N terminus) and the b ions (loss of residues at the C terminus) are labeled in the spectrum.

Table 3.

 

Effect of Various Inhibitors on Lens TOP Activity

Inhibitor	Relative Activity (%)*
		IQS-BK	Bradykinin
Control	100	100
TLCK (1 mM)	95	90
TPCK (1 mM)	85	97
EDTA (3 mM)	65	76
Phosphoramidon (1 mM)	85	90
PCMB (0.1 mM)	8	7
NEM (0.5 mM)	60	50
Bestatin (0.5 mM)	93	90
DFP (1 mM)	90	100
E64 (0.5 mM)	83	78
1,10-phenanthroline (0.1 mM)	5	4
Z-P-P-OH (0.5 mM)	95	90
Iodoacetamide (0.5 mM)	30	35
Biotin-RPPGF-cmk (0.1 mM)	40	0
Z-PheΨ(PO2CH2)-Ala-Arg-Phe (0.3 μM)	5	—
Pro-Ile (5 mM)	99	100
Lactacystin	90	92
CaCl2 (2 mM)	0	—
ZnCl2 (1 mM)	10	—
DTT (5 mM)	—	40

*  Enzyme activity was measured after 30 minutes of preincubation with various reagents at 25°C, using IQS-BK and bradykinin as substrates. β-Mercaptoethanol was excluded in the incubation buffer when thiol-reactive agents such as PCMB, NEM, and iodoacetamide were used.

Figure 7.

 

Effect of GSH on lens TOP activity. The purified protease was incubated with GSH in 20 mM Tris-HCl (pH 7.5) at 37°C. After 10 minutes, IQS-BK (3 nanomoles) was added, and the mixture was further incubated for 30 minutes at 37°C. The activity in the absence of GSH was taken as 100%.

Table 4.

 

Cleavage Sites in Peptide Substrates Used to Determine Bovine Lens TOP Specificity

Peptide Substrate	Cleavage Sites
Bradykinin	R-P-P-G-F+S-P-F-R
Neurotensin	pE-L-Y-E-N-K-P-R+R-P-Y-I-L
Adipokinetic hormone G	pE-V-N-F+S-T-G-W
IQS-BK	Mca-R-P-P-G-F+S-A-F-K (Dnp)

 The peptide substrates were incubated with purified lens enzyme for 30 minutes, and the cleavage products were separated by HPLC. The cleavage sites (+) were assigned by analyzing the products by MALDI-ToF mass spectrometry.

Table 5.

 

Relative Hydrolysis of Bradykinin Analogues by Bovine Lens TOP

Bradykinin Analogues (With C-Terminal Amidation)	Relative Activity*
(d)ARPPGFSPFR	100
(d)ARPPGYSPFR	380
(d)ARPPGRSPFR	108
(d)ARPPGLSPFR	260
(d)ARPPGDSPFR	0
(d)ARPPGFGPFR	110
(d)ARPPGFFPFR	82
(d)ARPPGFLPFR	8
(d)ARPPGFDPFR	0

 The bradykinin analogues were incubated with purified protease for 3 hours at 37°C, and the cleavage products were separated by HPLC. Standard peptides (d)ARPPGF and SPFR were used to calibrate the C18 reverse-phase column. The area under the peptide peaks in the HPLC profile was measured and taken as a measure of protease activity.

*  Enzyme activity with (d)ARPPGFSPFR substrate was taken as 100% for the purpose of calculation. The amino acids in bold contribute to the peptide bonds cleaved by TOP.

Figure 8.

 

Effect of ATP on TOP activity. (A) The dialyzed crude lens extract was preincubated with different ATP concentrations. The protease activity was measured using IQS-BK substrate in the presence of bestatin and Z-P-P-OH, as explained under methods. I, activity in presence of phosphinic peptide inhibitor. (B) Purified protease was incubated with 1 mM ATP for 1 hour. After this, bradykinin (BK) hydrolysis was monitored using a reverse-phase C18 column. An identical sample without ATP served as the control.

Figure 9.

 

Degradation of lens peptides by the purified TOP. The lens peptides were separated from old bovine lens extract. Purified protease (2 μg) was incubated with lens peptides for 1 hour at 37°C in 200 μL of 20 mM phosphate buffer (pH 7.0) containing 0.5 mM DTT and 0.05% Brij35. The concentration of newly formed amino groups was measured by the O-phthalaldehyde method. ²² The net proteolysis (A) is the difference between the concentration of amino groups before and after incubation with TOP. The TOP-specific degradation was confirmed by incubating the protease with a phosphinic inhibitor (I).

Table 6.

 

TOP Activity in Young and Aged Lenses

Species	Age (y)	IQS-BK Hydrolysis*
Bovine (n = 5 each)
Young	2–4	7.85 ± 0.30
Old	10–20	5.40 ± 0.48
Human (n = 3 each)
Young	18–25	3.70 ± 0.14
Old	60–80	3.26 ± 0.09

 TOP activity was measured using IQS-BK substrate. Z-P-P-OH, lactacystin and bestatin were included in the assay mixture to prevent the nonspecific hydrolysis of IQS-BK by other proteases present in the lens extracts.

*  Activity is expressed as nanomoles (mean ± SD) of substrate hydrolyzed in 1 minute per lens at 25°C.

The authors thank Beverly DaGue, Proteomics Center, University of Missouri, for mass spectrometry analysis.