Human lens crystallins are organized into three classes, α-, β-, and γ-crystallins, based on sequence homology and the size of aggregates isolated under physiologic conditions. Since lens crystallins undergo little or no turnover, there is opportunity for numerous post-translational modifications as the lens ages [1-4]. Some modifications of lens crystallins identified in young clear lenses may reflect normal development and maturation of the lens [5,6] whereas other modifications associated with aged lenses- such as deamidation, phosphorylation, truncation, glycation, oxidation, and cross-linking could negatively impact crystallin conformation, aggregation state, or solubility, resulting in increased light scattering and eventual loss of transparency in aged lenses [7-10].

In comparison to normal lenses, cataract lenses show greater fragmentation of crystallins due to proteolysis [1,11-13]. Increased crystallin fragmentation has also been reported in aged and cataract lenses of non-human species [6,14-17]. It has been reported that proteolysis may be a contributing factor in the insolubilization of crystallins occurring during normal maturation of lens or during cataract formation in human and bovine lenses [18]. It has been shown that the fragments of βA3/A1- and βB1-crystallins are selectively insolubilized during cataract development compared to normal aging [13]. These authors also reported increased crystallin truncation, the deamidation of Asn to Asp residues, and the oxidation of a Trp residue in cataract lenses.

Several proteolytic enzymes have been shown to play a role in aging of the lens and cataract formation [19-25]. Activities of several peptidases were reported to be highest in the outer cortical fibers and decreased to one half or below in the inner cortical fibers and nucleus. An inverse correlation between peptidase activities and the amount of crystallin fragments was observed in different regions of the lens. The amount of crystallin fragments and the amount of water-insoluble proteins were greater in the lens nucleus than in the outer cortical fibers [21]. An age-dependent decline in proteasome activity and a concomitant accumulation of modified proteins in human lens has been reported earlier [26]. Increased opacification of the lens nucleus in cataract was significantly correlated with decreased peptidase activities of the proteasome [27]. Although a protein oxidation reaction may render the protein susceptible to proteolysis, heavily oxidized proteins appear to first aggregate (via new hydrophobic and ionic bonds) and then to form covalent cross-links that make them highly resistant to proteolysis by 20S proteasome [28]. In fact, these aggregated, cross-linked oxidized proteins actually bind to the 20S proteasome and inhibit its ability to degrade the oxidized forms of other proteins [29].

Low molecular weight crystallin fragments have been isolated and characterized from both the water-soluble and water-insoluble fractions of the lens with greater prevalence in the water-insoluble fractions [11,21,30-32]. As many as 13 crystallin fragments with molecular masses between 3 and 17 kDa and originating from α-, β-, and γ-crystallins have been isolated from the water-soluble, high molecular weight aggregates of 60-80-year-old human lenses [31]. Accumulation of such crystallin fragments may be a cause for age-related lens opacity. It has been hypothesized that the interaction of short peptides derived from crystallins with lens proteins may increase the formation of high molecular weight aggregates and the scattering of light [33]. Though it is hypothesized that cataract develops as a result of improper interaction of crystallin fragments generated by proteolysis [25], the underlying mechanism is not clear. Earlier work by our group has shown that in vitro-oxidized crystallin peptides enhance the aggregation of β_L-crystallin and γ-crystallin and also exhibit anti-chaperone-like properties [33-35]. Interaction of peptide fragments with lens proteins and their aggregation-enhancing nature may have some implications for age-related cataract formation.

Recent studies in our laboratory have revealed the presence of 23 low molecular weight (<3.5 kDa) peptides in the urea-soluble fractions of normal young, normal aged, and aged cataract human lenses, and the amount of crystallin fragments was found to increase with age [36]. Out of these 23 peptides, 15 were derived from αA- and αB-crystallin. Most of the 15 came from the NH₂-terminal region, a few from the α-crystallin domain, and none from the COOH-terminal extension. Two crystallin peptides predominant in the nucleus of aged lens and cataract lens but not in young lens, αB_1–18 (MDIAIHHPWIRRPFFPFH, GenBank NP_001876) and βA3/A1_102–117 (SD(N)AYHIERLMSFRPIC, GenBank NP_005199), increased the molecular mass, polydispersity, and hydrodynamic radius of αA- and αB-crystallins [36]. Our study also showed that the peptides, αB_1–18 and βA3/A1_102–117, increased the scattering of light by β- and γ-crystallins and alcohol dehydrogenase (ADH) and also reduced the chaperone-like activity of α-crystallin.

The present study was undertaken to identify the binding sites for the βA3/A1_102–117 peptide in αB-crystallin. We used the photoactive cross-linker sulfo-SBED (sulfo-succinimidyl-2-[6-(biotinamido)-2-{p-azidobenzamido}-hexanoamido] ethyl-1–3 dithio propionate) [37] to confirm the interaction between αB-crystallin and the βA3/A1_102–117 peptide. The βA3/A1_102–117 peptide interacting sites in αB-crystallin are reported here.

The presence of crystallin fragments in water-soluble and water-insoluble fractions of lens proteins has been known for some years [11,40]. We first suggested that anti-chaperone peptides accumulating in vivo may be contributing to lens opacification during aging [33]. Recent studies in our laboratory have revealed the presence of 23 low molecular weight (<3.5 kDa) peptides in the urea-soluble fractions of normal young, normal aged, and aged cataract human lenses, and the amount of crystallin fragments was found to increase with age [36]. Our study also showed that βA3/A1_102–117 peptide, found in aged and cataract lens, increased the scattering of light by β- and γ-crystallins and alcohol dehydrogenase (ADH). βA3/A1_102–117 peptide also exhibited an anti-chaperone property by decreasing the ability of α-crystallin to prevent aggregation of β- and γ-crystallins and non-crystallin protein substrates, a process believed to be essential for maintenance of lens transparency [36].

The present study has confirmed the aggregation-enhancing property of the βA3/A1-crystallin-derived peptide, ¹⁰²SDAYHIERLMSFRPIC¹¹⁷, since it promoted the aggregation of β_L-crystallin during thermal denaturation (Figure 1A). The presence of this peptide in the assay mixture reduced the ability of αB-crystallin to protect β_L-crystallin against thermal aggregation (Figure 1B). Since the βA3/A1_102–117 peptide interacts with both β_L-crystallin and αB-crystallin, the enhanced light scattering observed in its presence may be due to a combination of its interaction with the substrate as well as the chaperone protein. To confirm the anti-chaperone nature of the peptide, we pretreated αB-crystallin with βA3/A1_102–117 peptide and separated the complex from the free peptide. The αB-crystallin-βA3/A1_102–117 peptide complex showed a considerable loss of chaperone-like activity against heat-denatured β_L-crystallin (Figure 2). Since free peptide was removed before the assay, we believe that the decreased chaperoning ability of αB-crystallin was specifically due to the binding of the peptide to αB-crystallin. When the βA3/A1_102–117 peptide was derivatized with sulfo-SBED, incubated with αB-crystallin, and photolyzed, the biotin label transferred to αB-crystallin (Figure 3, lane 1). The MALDI-TOF MS of the peptides derived from the αB-crystallin by trypsin digestion showed the presence of biotin from sulfo-SBED in several peptides (Figure 4). The identity of those peptides was obtained by sequencing in a nanospray QqTOF mass spectrometer. The sites on αB-crystallin where the label was observed (Figure 5) included the sequences, ⁷⁰LEKDR⁷⁴ and ⁸³HFSPEELKVK⁹². Earlier, we showed that ⁷³DRFSVNLDVKHFSPEELKVK⁹² is one of the well characterized functional chaperone sites in αB-crystallin [41]. Therefore, we can conclude that the anti-chaperone peptide βA3/A1_102–117 binds to αB-crystallin at the chaperone site. The binding of the βA3/A1_102–117 peptide to this substrate-binding site in αB-crystallin may have diminished the ability of the protein to effectively chaperone the denaturing β_L-crystallin.

Other sites of interaction between the βA3/A1_102–117 peptide and αB-crystallin included the sequences, ⁹¹VKVLGDVIEVHGKHEER¹⁰⁷ and ¹²¹KYR¹²³. These sequences are part of the α-crystallin domain. Earlier, Arg₁₂₀ was shown to be important for the chaperone-like activity of αB-crystallin [42]. A recent study demonstrated that ⁷³DRFSVNLDVKHFS⁸⁵ and ¹⁰¹HGKHEERQDE¹¹⁰ sequences in αB-crystallin inhibited the fibrillation of disease-related amyloidogenic proteins including α-synuclein and amyloid-β [43]. Other binding sites of the βA3/A1peptide to αB-crystallin included the COOH-terminal region of ¹⁵⁰KQVSGPER¹⁵⁷ and ¹⁶⁴EEKPAVTAAPKK¹⁷⁵. An earlier study has shown that mutations of the two terminal lysines (174 and 175) of αB-crystallin to leucine or glycine greatly reduced its chaperone-like activity. It was postulated that mutations in these two amino acids may have resulted in the COOH-terminus folding back on itself, losing its flexibility, and sterically hindering protein binding [44]. The COOH-terminal extensions of α-crystallins are considered polar, highly flexible, and solvent-exposed. These extensions also act as solubilizing agents for the relatively hydrophobic α-crystallin molecule and the high molecular weight complex that forms during the chaperone action [45].

αB-Crystallin with five amino acids deleted from its COOH-terminus was found in lenses of hereditary cataractous iCR/f rat, and this truncated crystallin showed decreased chaperone activity [46]. Protein pin array studies have identified ¹⁵⁷RTIPITRE¹⁶⁴ as one of the interactive sequences for chaperone activity in human αB-crystallin [47]. Deletion of the polar COOH-terminal sequence, ¹⁵⁵PERTIPITREE¹⁶⁵, from human αB-crystallin resulted in poor solubility and limited or no chaperone activity against unfolding β_L-crystallin, alcohol dehydrogenase, and citrate synthase. These results demonstrated the importance of the COOH-terminal residues in the recognition, selection, and maintenance of solubility of unfolding substrate proteins [48]. As identified by protein pin-array studies, the ¹⁵⁰KQVSGPER¹⁵⁷, ¹⁶⁴EEKPAVTAAPK¹⁷⁴, and ¹⁶⁴EEKPAVTAAPKK¹⁷⁵ sequences, which are identified as peptide interacting sites in αB-crystallin in the present study, overlap with sequences involved in dimerization and assembly of human αB-crystallin [49]. The binding of the βA3/A1_102–117 peptide to the COOH-terminal extension of αB-crystallin in the present study may have interfered with substrate binding and decreased the chaperoning ability of the protein. Our recent studies showed that the βA3/A1_102–117 peptide caused partial precipitation of αB-crystallin [36]. We believe that the peptide interaction at the COOH-terminal extension of αB-crystallin may have led to decreased flexibility and decreased solubilization potential of αB-crystallin. Additionally, the interactions between crystallins and low molecular weight peptides may limit the access of these peptides to peptidases, resulting in the accumulation of peptides with aging.

The substituted peptide we used in this study, SDADHGERLMSFRPIC, had two hydrophobic residues (tyrosine and isoleucine) replaced by hydrophilic (aspartate and glycine) residues. The substitution abolished the ability of the peptide to induce β_L-crystallin aggregation (Figure 1A) and to act as anti-chaperone (Figure 1B). This suggests that the hydrophobic amino acids play a critical role in the interaction of the anti-chaperone peptide with αB-crystallin. The substituted peptide can also be considered a control peptide for the label transfer study. If there were any non-specific interactions between the derivatized substituted peptide and αB-crystallin, we would have seen a transfer of biotin to αB-crystallin following photolysis. The fact that the αB-crystallin incubated with derivatized substituted peptide showed only about 1.2% of the label transfer when compared to the derivatized βA3/A1_102–117 peptide during the blotting experiments (Figure 3 lane 2) suggests that the non-specific interaction and label transfer was insignificant in this study.

In summary, this study confirmed the interactions between βA3/A1_102–117 peptide and αB-crystallin. The binding of this peptide to regions critical for chaperone action of αB-crystallin may explain its anti-chaperone property. The modulation of structural and functional properties of crystallins by low molecular weight peptides may play an important role in protein aggregation and cataract formation.