Effect of Deamidation of Asparagine 146 on Functional and Structural Properties of Human Lens αB-Crystallin

Abstract

purpose. To elucidate the effect of deamidation on the structural and functional properties of human αB-crystallin.

methods. Site-directed mutagenesis was used to generate three deamidated mutants of αB-crystallin: N78D, N146D, and N78D/N146D. The mutations were confirmed by DNA sequencing and matrix-assisted desorption ionization–time of flight (MALDI-TOF) mass spectrometry. Recombinant native αB-crystallin (wild type [WT]) and the three mutated αB species were expressed, and each species was purified to homogeneity by ion-exchange chromatography followed by hydrophobic interaction chromatography. The structural and functional properties compared with WT protein were investigated, respectively, by static light scattering (SLS), circular dichroism (CD), and fluorescence spectroscopy and by determining chaperone activity with the use of three substrates.

results. Native WT and the N78D mutant showed relatively higher chaperone activity compared with the N146D and N78D/N146D mutants with all the substrates. Further, during binding experiments with 1-anilino-8-naphthalenesulfonate (ANS), the WT and N78D mutant showed relatively more solvent-exposed hydrophobic residues than the N146D and N78D/N146D mutants. On determining far-UV circular dichroism and tryptophan (Trp) fluorescence spectra, significant secondary and tertiary structural changes were observed in the N146D and N78D/N146D mutants compared with WT and the N78D mutant. The static light scattering data showed a high order of oligomerization in all the three mutants. N146D and N78D/N146D formed the largest oligomers of 750 and 770 kDa, respectively, compared with WT (580 kDa).

conclusions. The results show that the deamidation of N₁₄₆ but not of N₇₈ have profound effects on the structural and functional properties of αB-crystallin.

Crystallins are the structural proteins of mammalian lens and belong to two classes, the α-crystallin and the βγ-crystallin superfamilies. α-Crystallin is one of the most abundant proteins in lens mature fiber cells, constituting approximately 35% of total lens proteins, and exists in vivo as an 800-kDa oligomer of two subunits of αA- and αB-crystallins at a ratio of 3:1. ^{1 2}  

Besides being a structural protein, αB-crystallin is a small heat shock protein (sHSP), ² and is structurally related to other sHSPs including αA-crystallin, hsp27, hsp20, hsp22, and myotonic dystrophy protein kinase binding protein. ³ It shares 57% homology with αA-crystallin and plays extralenticular roles in both normal and diseased tissues. ⁴ Reports show an overexpression of αB-crystallin in the development of benign tumors associated with tuberous sclerosis, neuromuscular disorders, ¹ and other neurologic diseases, such as Alexander’s, Alzheimer’s, and Parkinson’s diseases. ^{1 5}  

Bovine α-crystallin in a dilute solution at 37°C had a molecular weight (M_r) of approximately 550,000, and its M_r in its native state was approximately 700,000. ^{6 7} A recent determination of molar mass distribution by a light-scattering instrument showed polydispersity in α-crystallin, with the average M_r of 700,000 for bovine α-crystallin and 585,000 for human recombinant αB-crystallin. ⁵ Because the accuracy of the light-scattering system was 3% to 5%, the M_r was fairly accurate. 

It is believed that oligomerization or multimeric assembly is a prerequisite for the molecular chaperone activity of αA- and αB-crystallins, which prevents undesirable protein interactions and their aggregation, and therefore contributes to the maintenance of lens integrity and transparency. ^{4 5 8} This chaperone activity of the crystallins is known to decrease with age and during cataractogenesis. ⁹ According to Horwitz, ⁵ although the questions about the ability of α-crystallin to refold denatured protein ⁵ remains, several examples show its ability to protect enzymes from denaturation. ^{10 11}  

Cataractogenesis (development of lens opacity) is believed to be a consequence of accumulation of insoluble aggregates and cross-linked products of α-, β-, and γ-crystallins. The insolubilization of crystallins is thought to be initiated by posttranslational modifications, ^{12 13 14} which change their structural and functional properties. During aging and cataract formation, αB-crystallin undergoes numerous in vivo posttranslational modifications, such as phosphorylation of amino acid S; N-terminal acetylation; oxidation of M and C residues; degradation and deamidation of N and Q; and carbamylation of K and R residues. ^{12 13 15 16 17} It is speculated that these modifications either act independently or in combination to cause unfolding, aggregation, or cross-linking of proteins, which in turn leads to protein insolubilization and to lens opacity. ^{12 13} However, modification-induced mechanisms are poorly understood. 

Deamidation of crystallins is a nonenzymatic process and is one of the most prevalent posttranslational modifications that occur during aging and cataract formation. ^{1 2} Both N and Q amino acids have been shown to undergo deamidation, ¹² which results in the introduction of negative charge at the site of modification and thus may alter the protein’s tertiary structure and affect its biological properties. Because deamidation changes the native conformation, it is believed to change the protein’s three-dimensional structure. ^{12 13} This has been shown in several studies; however, we still poorly understand the exact role of deamidation in protein destabilization during aging and cataractogenesis and have not identified a definitive cause-and-effect relationship between deamidation and chaperone activity. Recent studies have shown a correlation between two posttranslational modifications (oxidation and phosphorylation) and the chaperone activity of αB-crystallin. ^{9 18 19} However, presently, no correlative study exists in the literature that shows the effect of deamidation on the chaperone activity of αA- and αB-crystallins. 

The in vivo deamidation of α-, β-, and γ-crystallins has been shown in several studies. ^{12 13 15 16 20 21 22} A recent study from our laboratory has shown deamidation of N₁₄₆ in αB-crystallin of normal and cataractous human lenses. ¹⁶ Similarly, the increased deamidation of N₁₄₃ in γS-crystallin ^{20 22} and N₁₀₁ in human αA-crystallin and their aggregation in high molecular weight species in normal human lenses have been reported. ²¹ In addition, Hanson et al., ¹² have reported that deamidation is significantly greater than other modifications in crystallins in lenses from 50- to 60-year-old donors. Similarly, Lampi et al., ²³ have shown that deamidation and truncation of human βB1-crystallin lead to structural destabilization and decreased stability in higher temperatures. In a later study, they reported that deamidation alters the structure of the βB1 dimer. ²⁴  

As stated, we identified deamidation of N₁₄₆ in a fragment of human αB-crystallin isolated from normal and cataractous lenses of 60- to 80-year-old donors. ¹⁶ Such deamidation of N₁₄₆ has also been reported in bovine αB-crystallin, ¹⁵ but the effect of deamidation on structure and functional properties of the crystallin is presently unknown. One way to study this is to use site-specific mutagenesis to generate deamidated αB-crystallin mutants and determine their structural and functional properties. Therefore, in this study, these mutants were used to determine the effects of deamidation on conformation, the microenvironment of various amino acids, and the chaperone activity of αB-crystallin, to gain an insight into the structure and functional properties of the crystallin. 

In the present study, three mutant forms of human αB-crystallin (N78D, N146D, N78D/N146D) were generated and comparative analyses were performed to determine their chaperone activities and structural properties by static light-scattering (SLS), circular dichroism (CD), and fluorescence spectroscopy. The results show that deamidation at N₇₈ in αB-crystallin has relatively less effect on functional and structural properties than does deamidation at N₁₄₆. 

Materials and Methods

The restriction endonucleases NcoI and EcoRV and the molecular weight protein markers and DNA markers were purchased from Amersham Biosciences (Arlington Heights, IL) and Promega (Madison, WI), respectively. T7 promoter, T7 terminator, and other primers used in the study were obtained either from Invitrogen (Carlsbad, CA) or from the University of Alabama at Birmingham’s Oligo Synthesizing Core Facility. Molecular biology–grade chemicals were purchased from Sigma-Aldrich (St. Louis, MO), unless otherwise stated. All chemicals for two-dimensional (2-D) gel electrophoresis were from Amersham Biosciences or BioRad (Hercules, CA). Diethylaminoethyl (DEAE)-Sephacel and butyl-Sepharose were obtained from Amersham Biosciences. 

Bacterial Strains and Plasmids

The bacterial strain Escherichia coli BL21 (DE3)pLysS was obtained from Promega. The human αB-crystallin cDNA cloned on a plasmid pDIRECT was received from Mark Petrash, Washington University (St. Louis, MO). Cells were propagated in Luria broth, and recombinant bacteria were selected with ampicillin. 

Site-Specific Mutagenesis

Deamidation of the N residue to the D residue at positions 78, 146, or both in αB-cDNA was induced with a site-directed mutagenesis kit, used according to the manufacturer’s instructions (Quickchange; Stratagene, La Jolla, CA). The recombinant human αB-crystallin coding sequence was cloned in pDIRECT, as described earlier. ²⁵ These constructs were used as templates with the mutated primers (where N was replaced by D, Table 1 ) to generate site-specific mutations. Briefly, 25 ng of template was used, and the PCR conditions were as follows: predenaturing at 95°C for 30 seconds, followed by 16 cycles of denaturing at 95°C for 30 seconds, annealing at 55°C for 1 minute, and extensions at 68°C for 5 minutes. After digestion with DpnI, 2 μL of the PCR product was used to transform the XL1-Blue supercompetent cells provided with the kit (Quickchange; Stratagene). The sequencing of desired DNA preparation at the University of Alabama at Birmingham’s Core Facility identified the positive clones. 

Expression and Purification of Wild-Type and Deamidated Proteins

E. coli BL21 (DE3)pLysS was transformed with mutant amplicons according to a standard procedure. ²⁶ The proteins were overexpressed by addition of isopropylthiogalactoside (IPTG; final concentration of 1 mM) and the cultures were incubated further at 37°C for 3 hours. Bacterial cells were lysed in 50 mM Tris-HCl buffer (pH 7.5) containing 0.5 mM EDTA and 0.3 mM NaCl (TEN buffer). The isolation and purification of the mutant proteins was performed as described by Merck et al. ²⁷ and Pasta et al., ²⁸ with a few modifications, as follows. The soluble fraction was dialyzed against 50 mM Tris-HCl buffer (pH 7.9) containing 0.5 mM EDTA and 1 mM dithiothreitol (DTT; TED buffer) and subjected to DEAE-Sephacel ion-exchange chromatography. The bound proteins were eluted with a gradient of 0 to 0.5 M NaCl in TED buffer. Next, the fractions containing αB-crystallin were identified by SDS-PAGE (15%), pooled, and subjected to hydrophobic interaction chromatography using a butyl-Sepharose column. During the chromatography, the equilibration buffer contained 50 mM phosphate buffer (pH 7.0) containing 0.5 M (NH₄)₂SO₄, and the bound proteins were eluted with a decreasing (NH₄)₂SO₄ concentration (i.e., 0.5–0.0 M). Again, the αB-crystallin–containing fractions were analyzed by SDS-PAGE, concentrated by lyophilization, and dialyzed against water at 5°C for 24 hours. The purity of the wild-type (WT) and mutant αB-crystallin species was examined by SDS-PAGE and 2-D gel electrophoresis. The concentrations of the proteins were determined by using a protein determination kit (Pierce, Rockford, IL) and by absorbance at 280 nm, as required. 

2-D Gel Electrophoresis

The protein samples were mixed with resolubilization buffer (5 M urea, 2 M thiourea, 2% 3-[C3-cholamidopropyl]dimethtyl-ammonio-1-propansulfonate (CHAPS), 2% caprylylsulfobetaine 3-10, 2 mM tri-butyl phosphine, and 40 mM Tris [pH 8.0]) in a 2:1 ratio. ²⁹ Each sample was subjected to 2-D gel electrophoresis (isoelectric focusing [IEF] in the first dimension and SDS-PAGE in the second dimension). IEF separation was performed (Immobiline DryStrips; pH range, 3.0–10.0), by following the manufacturer’s instructions (Amersham Biosciences). SDS-PAGE in the second dimension was performed with a 15% polyacrylamide gel. ³⁰  

Chaperone Activity Assays

The chaperone activity of the WT and the deamidated αB-crystallin species was studied by using three assay methods, in which the aggregation of the target protein was induced either by thermal or nonthermal means. In these analyses, the aggregation of the target proteins was monitored at 360 nm (due to light-scattering) as a function of time (60 minutes) using a scanning spectrophotometer (model UV2101 PC; Shimadzu, Columbia, MD) equipped with a six-cell positioner (model CPS-260; Shimadzu) and a temperature controller (model CPS 260; Shimadzu). With insulin as a substrate, ³¹ the aggregation was determined at 25°C. Aggregation of insulin (1 μM in 10 mM phosphate buffer [pH 7.4] containing 100 mM NaCl) was initiated by 20 mM DTT at different chaperone to target protein molar ratios (αB-crystallin to insulin) as described in the Results section. During the thermal aggregation assay, 1 μM citrate synthase (CS, in 50 mM phosphate buffer [pH 7.8] containing 150 mM NaCl and 2 mM EDTA) was incubated at 43°C with various concentrations of αB species to obtain the desired chaperone-to-target protein molar ratios. ³² In the third assay, aggregation of human γD-crystallin was monitored at 63°C. ⁴ Human recombinant γD-crystallin was purified as previously described by Srivastava and Srivastava. ³³ Recombinant human γD-crystallin (1 μM) in phosphate buffer was incubated in the absence or presence of various concentrations of the αB species, and the identical assay was performed for each concentration. 

Fluorescence Studies

All fluorescence spectra were recorded in corrected spectrum mode with a spectrofluorometer (RF-5301PC; Shimadzu) with excitation and emission band-passes set at 5 and 3 nm, respectively. The total tryptophan (Trp) intensities of the WT (0.15 mg/mL) or the deamidated αB species (0.15 mg/mL) in 10 mM phosphate buffer (pH 7.4) containing 100 mM NaCl were recorded by excitation at 295 nm. 

The binding of a hydrophobic probe, 8-anilino-1-naphthalenesulfate (ANS) to WT and αB mutants was determined as described previously. ²⁷ To the WT or deamidated αB species (0.15 mg/mL) in 10 mM phosphate buffer (pH 7.4) containing 100 mM NaCl, 15 μL of 0.8 mM methanolic ANS solution was added and the preparation incubated for 10 minutes at 37°C. Fluorescence spectra of the samples were recorded from 400 to 600 nm with excitation at 390 nm. In another experiment, the WT and mutated protein preparations were heated with the same concentration of ANS at 43°C for 10 minutes and fluorescence spectra determined after cooling the samples to room temperature. 

Circular Dichroism Studies

Far-UV CD spectra of the WT and deamidated αB species were recorded with a CD spectropolarimeter (Model 62 DS; Aviv, Lakewood, NJ). Experiments were performed with 1.0 mg/mL of protein in 50 mM potassium phosphate buffer (pH 7.4), using a 0.1-cm-path-length cell for the far-UV region. The protein concentrations were determined based on the absorption at 280 nm from the absorption spectra of desired proteins between 230 and 300 nm. All spectra reported are the average of five accumulations. 

Static Light-Scattering

The light-scattering experiments were performed on an SLS instrument (Model 202; Precision Detectors, Bellingham, MA). Protein samples in 50 mM Tris-HCl (pH 7.9) were filtered through a 0.2-μm filter before analysis. Results used both 90° and 15° light-scattering detection. 

Results

Confirmation of Site-Specific Mutations in αB-Crystallin Mutants

Human lens αB-Crystallin contains two asparagines (N) at positions 78 and 146. By point mutation, each of the N residues was changed to D in two individual mutants, and both N₇₈ and N₁₄₆ were changed to D in one mutant. The resultant deamidated proteins are referred to as N78D (N at position 78 changed to D), N146D (N at position 146 changed to D), and N78D/N146D (N at 78/146 positions changed to D) mutants throughout the text. 

DNA sequencing data confirmed the mutations at the desired positions in the three mutants (results not shown). To confirm this further, the expressed WT and mutant proteins after trypsin digestion were analyzed by matrix-assisted desorption ionization–time of flight (MALDI-TOF) mass spectrometry. The isotopic distribution of tryptic fragments of αB-crystallin species further confirmed mutation(s) of N to D residue at the desired positions. A tryptic peptide with a mass of 921.57 was detected in WT representing residue 75-82 (sequence: RFSVNLDVK) of αB-crystallin (Fig. 1A) . In the mutant N78D, the major peak observed had a mass of 922.56, suggesting a gain of 1 mass unit, due to mutation of N₇₈ to D₇₈ (Fig. 1C) . Similarly, a peptide with a mass of 2625, representing residue 124-149 (sequence: RIPADVDPLTITSSLSSDGVLTVNGPR) of αB-crystallin was seen in the WT (Fig. 1B) , but a peptide with mass of 2626.08 was relatively higher in the mutant N146D compared with a peptide with mass of 2625 (Fig. 1D) . A gain of 1 mass unit in the peptide with a mass of 2625 in the mutant compared with WT suggested mutation of N₁₄₆ to D₁₄₆. Similar results (i.e., gain of 1 mass unit) were obtained in two peptides with a mass of 922.56 and 2626.08 from the N78D/N146D mutant compared with the WT, confirming the mutations at both the sites (results not shown). 

Expression and Purification of Human Recombinant αB- and Deamidated αB-Crystallin Mutants

The WT and the mutated αB-species were expressed in E. coli. SDS-PAGE analysis and MALDI-TOF analyses showed the expression of full-length recombinant αB-crystallin in the system used. The expressed proteins were purified to homogeneity as described in the experimental procedure in the Methods section, using ion-exchange and hydrophobic chromatography. After purification, SDS-PAGE analysis showed a single protein band of 20,000 of the WT and three deamidated αB mutant species (Fig. 2) . During the hydrophobic chromatography, all three mutants were found to be more acidic than the WT species. The observation suggested that the deamidated mutants differ considerably from the WT in columbic interactions. 

Comparison of the Effect of Deamidation on the Chaperone Activity of WT and Deamidated αB Mutants

As stated earlier, the chaperone assay was performed using three different substrates at room temperature or at elevated temperatures. Figure 3A , shows chaperone activity of αB-crystallin species on DTT-induced aggregation of insulin at 25°C at various chaperonin-to-insulin ratios. The WT showed a maximum chaperone activity at all the ratios of chaperonin to insulin used. The percentage protection from aggregation (monitored at 360 nm) of insulin increased as the concentrations of WT αB-crystallin were increased. At the 1.2:1.2 molar ratio of insulin to WT αB or mutant, the WT showed 50% protection, whereas the N78D showed 40% and N146D and N78D/N146D mutants showed only 8% to 20% protection. At a higher ratio of chaperonin to insulin (2:1), the activity of the WT increased to 80% and of the N78D mutant to 45%, but it remained only 8% to 10% in the N146D and N78D/N146D mutants. 

Figure 3B shows the chaperone activity of WT and the three αB-mutants against heat-induced aggregation of CS at 43°C. Again, the WT showed increasing chaperone activity with increase in concentration. Further, whereas the N78D mutant showed almost the same level of chaperone activity as the WT, the N146D and N78D/N146D mutants showed very little chaperone activity (5%–8%) compared with WT and the N78D mutant at all the ratios. 

Figure 3C shows the chaperone activity of WT and deamidated αB mutants during heat-induced denaturation at 63°C of human recombinant γD-crystallin at different molar ratios of γD and WT or mutant proteins. The WT showed maximum chaperone activity compared with the three mutants; however, the protection of denaturation of γD-crystallin was only 40% compared with the activity observed during the two assays. All three mutants showed very little chaperone activity at the two ratios of 0.3:1 and 0.6:1 (chaperonin to γD-crystallin), but, at the higher molar ratio of 1.2:1.2, the mutant N78D showed relatively higher chaperone activity compared with the other two (N146D and N78D/N146D) mutants. Together, the results show that although the chaperone activity of WT and the three deamidated αB mutants varied in the three assay systems, typically the order of chaperone activity remained the same (i.e., WT>N78D> N146D≥N78D/N146D). Further, the chaperone activity of N146D and N78D/N146D mutants was considerably reduced compared with WT and N78D, suggesting the relatively greater role of N₁₄₆ compared with N₇₈ in the maintenance of chaperone function of αB-crystallin. These findings are supported by the earlier reports, which showed that the difference in stoichiometry of chaperone to target protein affects chaperone activity. ^{4 28}  

Surface Hydrophobicity of WT and αB Deamidated Mutants

ANS (a hydrophobic probe) is nonfluorescent in aqueous solution and becomes fluorescent when bound to the hydrophobic residues on the surface of a molecule. Figure 4A shows the fluorescence intensity of WT and the mutant proteins after binding with ANS at 37°C. At 12 μM ANS concentration, the fluorescence intensity of ANS bound to WT was at its maximum compared with the mutants. At an identical concentration, the fluorescence intensity of ANS-bound to N78D, N146D, and N78D/N146D was 85%, 60%, and 25%, respectively, compared with WT. The data suggest that after mutation of the N₁₄₆ residue, the hydrophobicity of the mutants compared with WT was significantly reduced, but similar deamidation of N₇₈ caused little change in hydrophobicity. 

Figure 4B shows the surface hydrophobicity of WT and mutant proteins after ANS binding at 43°C. In this experiment, the samples were heated to 43°C with ANS and the fluorescence intensity of each sample was measured after cooling to room temperature. The fluorescence intensity of ANS bound to WT and N78D at 43°C was 10% higher than the intensity observed at the physiological temperature (37°C). In contrast, the fluorescence intensity of ANS bound to the mutants N146D and N78D/N146D at 43°C decreased significantly compared with ANS-binding at the physiological temperature (37°C). Together, the data suggest that the relative surface hydrophobicity of WT and mutants were dependent on temperature, but in general the hydrophobicity of the N146D and N78D/N146D mutants was lower compared with the WT and the N78D mutant. 

Secondary and Tertiary Structure Analyses

Comparative analyses of secondary and tertiary structure of WT and the three deamidated αB-mutants were performed by determining their CD spectra and Trp fluorescence, respectively. Tryptophan fluorescence data reflect a protein’s gross positioning of Trp residues. Proteins that differ in the exposure of their Trp residues may differ in the wavelength of illumination at which fluorescence emission maxima are obtained (λ_max). In our study, the fluorescence emission spectrum of WT had the highest fluorescence intensity and a λ_max at 338 nm (Fig. 5) . The N78D mutant showed 50% fluorescence intensity and an emission maximum at 340 nm, indicating an insignificant change in its local environment of the Trp residue compared with WT. The N146D and N78D/N146D mutants showed an emission maximum at 330 and 332 nm, respectively, the Trp fluorescence was highly quenched in these two mutants compared with the WT, suggesting substantial changes in the microenvironment around the Trp in the N146 mutant. 

Far UV CD of WT and the Deamidated Mutants

The changes in secondary structure in WT and the three mutants were examined by determining their far-UV CD spectra. The variation of CD as a function of wavelength is shown in Figure 6 . The CD spectrum of WT showed a minimum around 210 to 212 nm, unlike the typical 218-nm band for β-sheet conformation. The results presented herein are similar to those in a previous study, which showed a difference in λ_max in the negative band in αA- and αB-crystallins. ³⁴ The present study used the PROSEC program to deconvolute the CD spectra. The estimates of secondary structure obtained indicated that the conformation of the WT consisted primarily of β-sheet (58%–65%) with some α-helix, which is consistent with the previously published data. ³⁴ No significant change in CD spectra was observed in the N78D mutant compared with the N146D and N78D/N146D mutants (Fig. 6) . Although the intensity of the negative ellipticity did not significantly change in the case of the mutants N146D and N78D/N146D, a shift in λ_max in comparison to WT was observed. The mutants showed maximum ellipticity at 207 to 210 nm, indicating the occurrence of conformational transitions. All the three mutants showed decreased β-sheet content (both β-sheet and β-turn combined were 55%–60%) compared with WT. Because the native αB-crystallin has predominantly β-sheet conformation, the induced conformational changes appear to be a consequence of deamidation of the N residue at position 146 in the mutant. 

Quaternary Structure of WT and the Deamidated Mutants

To compare the quaternary structure of the WT and the mutants, the molecular weights of these species were determined by an SLS instrument as described in the experimental procedures. The method is reliable because it uses gel-permeation chromatography coupled to low- and high-angle laser light-scattering and a differential refractive index detector to determine the molecular weight. Figure 7 shows the elution profile and the molecular mass of WT and the three mutants. Based on the difference of elution of WT and the mutants from an SLS column (TSK G-4000 PW_XL; Toso Haas, Montgomeryville, PA) and their size detection by the light-scattering instrument, the sizes of the oligomeric proteins were determined. The mutant N78D/N146D (770,000) eluted last followed by N146D (750,000) and N78D (670,000) just ahead of WT αB-crystallin (580,000). Together, the data suggest that the mutants formed larger oligomers than the native αB-crystallin (Fig. 7) . The molecular weight at the peak obtained from the light-scattering measurements is shown in Table 2 . The observed molecular weight of αB-crystallin oligomer was found to be 580,000, which matched with an earlier reported study. ⁵  

Discussion

In a previous study, ¹⁶ we showed that N₁₄₆ residue in human αB-crystallin undergoes in vivo deamidation, and several fragments containing this modification were found in both water-soluble and -insoluble protein fractions of normal and cataractous human lenses. The fact that only the deamidation of N₁₄₆ residue was the major posttranslational modification in the αB fragments suggested that the deamidation may cause structural changes in αB-crystallin, leading to water insolubilization. The deamidation of crystallins increases with age, ^{12 21 22} and it is more prevalent in crystallins present in the water-insoluble fraction. ^{12 13} Further, deamidation of N₁₄₃ in γS-crystallin has been shown to be a cataract-specific but not an age-specific event. ²¹ Because the specific effects of deamidation on structural and functional properties of αB-crystallin and potential roles in development of cataract is unknown, the present study was undertaken. 

The human lens αB-crystallin contains two N residues at positions 78 and 146, and these were mutated to D with the purpose of determining the effect of deamidation on the structural and functional properties of αB-crystallin. Because this is the first study to initiate structural changes by introduction of a negative charge in αB-crystallin and the deamidation is reported to be a most common modification in lens crystallins, the present study is highly relevant in understanding the potential role of deamidation in cataractogenesis. Further, the deamidated αB-crystallin mutants allowed us to correlate the molecular effects on functional (chaperone activity) and the structural properties. 

A major finding of this study was that the deamidation of the N₁₄₆ residue and not the N₇₈ residue had a profound effect on the chaperone activity and structural properties of αB-crystallin. The conclusion about the functional property was based on the determination of the chaperone activity of the WT and deamidated αB-crystallin mutants by three different assay systems with three different substrates: DTT-induced aggregation of insulin at room temperature, and denaturation of CS and γD-crystallin at 43°C and 63°C, respectively. These assays provided almost identical results. The three assay systems were chosen because previous studies have shown variations in the chaperone activity of αA and αB-crystallins based on the type of assay system used. ^{4 28}  

The chaperone activity of N78D and N146D varied in the three assays. During the assay with CS, the N78D mutant showed almost the same level of chaperone activity as did the WT at four stoichiometric ratios of CS to αB species. With this system, the N146D and N78D/N146D mutant showed almost no chaperone activity. During the assay with insulin at various stoichiometric ratios, again the N146D and N78D/N146D mutants showed little chaperone activity (i.e., only 10%–20% of total WT activity), but the N78D mutant had almost 50% to 75% of the activity observed with the WT. In the assay with γD-crystallin, all three mutants showed very little chaperone activity. In a comparison of the chaperone activity at a 1.2:1.2 ratio (substrate to WT or mutants of the αB species) during the three assays, the order of increasing activity was identical (WT>N78D> N146D≥N78D/N146D). 

The nearly equal levels of chaperone activity in the N78D mutant and the WT suggest that the alteration in a surface charge due to deamidation at N₇₈ caused little conformational change or surface exposure compared with the other two mutants. A recent study of human γS-crystallin has shown that deamidation is influenced by surface exposure (i.e., minimal deamidation for residues with accessibility number 8 nm). ³⁵ Therefore, the relatively lower rate of deamidation at N₇₈ in comparison to N₁₄₆ in αB-crystallin is probably due to relatively lower accessibility of the N₇₈ residue than of the N₁₄₆ residue during deamidation. As stated earlier, in a recent study we found deamidation at N₁₄₆ but not at N₇₈ in αB fragments. ¹⁶ Together, apparently the deamidation at N₇₈ in αB-crystallin is not naturally favorable, and as described earlier, the deamidation at this position only moderately alters the chaperone function of the crystallin. 

Interaction between the chaperone activity and the target binding sites involves hydrophobic patches in α-crystallin, but these are not the sole determinant. ^{36 37 38} A previous report has shown that at physiological temperature (37°C), αA- and αB homopolymers show almost the same levels of hydrophobicity. However, in another study, αA-crystallin was shown to be a better chaperone at higher temperature, due to conformational changes that exposed additional hydrophobic sites, whereas no such transition occurred in αB-crystallin. ³⁶  

To gain an insight into the differences in chaperone activity of the WT and mutants of αB-crystallin, we compared structural properties of these species. A monomeric dye, ANS, is known to bind to solvent-exposed hydrophobic residues and become highly fluorescent. Both bis-ANS and ANS have been used to probe hydrophobicity of αA- and αB-crystallins. ³⁶ The fluorescence spectra of the probe bound to WT and the N78D mutant showed a moderate increase (10%) at 43°C compared with that shown at physiological temperature (37°C). However, N146D and N78D/N146D mutants showed a significant decrease in the fluorescence intensity at 43°C compared with that at 37°C, indicating lesser accessible hydrophobic residues at the elevated temperature. This finding was further supported by the results of intrinsic Trp fluorescence spectra, which showed a decrease in maximum emission in the above two mutants compared with WT. Together, the data suggest exposure of a relatively lesser number of hydrophobic residues in the mutants compared with WT, which may also explain the lower chaperone activity in the two mutants at 43°C with CS as a substrate. Apparently, when the temperature was increased from 37°C to 43°C, the two deamidated mutants (N146D and N78D/N146D) underwent conformational changes that led to burying of certain hydrophobic residues and in turn a reduction in the sites available for substrate binding during chaperone activity. In contrast, the deamidation of N₇₈ in the αB mutant had minimal effect on chaperone activity because of the hydrophobic residues had same level of exposure as in the WT. 

The fluorescence characteristics of a Trp residue are dependent on its microenvironment. αB-crystallin contains two Trp at positions 9 and 60. The local environment of Trp was examined by the intrinsic Trp fluorescence. The fluorescence emission varies from 320 nm in an apolar solvent to 350 nm in water. The λ_max therefore throws some light on the polarity of the Trp residues—that is, the greater the emission the higher the levels of free Trp residue accessible in water. Because the N146D and N78D/N146D mutants showed a λ_max at 330 to 332 nm, which is lower than the WT (339 nm), apparently the deamidation at N₁₄₆ changed the microenvironment around Trp in the two mutants compared with WT. However, the exact organization of these aromatic residues will be known once the crystallographic structure of αB-crystallin is available. 

To determine the cause of reduced chaperone function in the mutants compared with WT, the structural changes because of the mutation were investigated. The far UV CD spectra revealed that the structure of αB-crystallin was not affected on deamidation at N₇₈ because negative ellipticity at 210 to 212 nm was observed, although the intensity was lower compared with that in WT. The N146D and N78D/N146D mutants showed a negative band at 207 to 210 nm compared with the 210- to 212-nm bands for WT, indicating the induction of helical conformation. Because αB-crystallin has a predominantly β-conformation, the results suggests that the conformational transition may be induced by deamidation at N₁₄₆. 

Previous studies of CD spectra ³⁹ and Fourier transform infrared measurements ³⁶ have shown that the secondary structures of both αA- and αB homopolymers are similar, with a slightly higher content of a β-sheet structure (and lower proportion of α-helix) in αB-crystallin. This study ³⁶ also concluded that the thermotropic changes in the secondary structures of αA- and αB-crystallins were identical and could not account for the heat-induced increase in the chaperone activity in αA-crystallin. In contrast, as shown earlier, differences in the CD spectra of the WT and mutant αB species were seen in this study. This suggests that deamidation has a profound effect on the structure of αB-crystallin. 

Our studies of hydrophobic site-binding with ANS, the Trp fluorescence, and far UV-CD spectra indicate the introduction of a negative charge after deamidation at N₁₄₆ alters the secondary structure and results in loss of chaperone function. Because αB-crystallin exists as a multimer, and it has been shown that oligomerization is a prerequisite for chaperone activity, quaternary structure of the deamidated mutants was compared with the WT. The SLS data showed that the N146D and N78D/N146D mutants formed the largest oligomer of 750,000 and 770,000, respectively, compared with N78D (670,000) and WT (580,000). The alterations in secondary structures also caused changes in the oligomerization property of the αB-mutants. As stated earlier, the mutants with relatively higher oligomers also exhibited the lower chaperone activity, lower fluorescence intensity, and lower hydrophobicity (i.e., Trp spectra and ANS binding) compared with WT, indicating that introduction of negative charge on deamidation at N₁₄₆ results in an inefficient packing (loosely organized) of the structure, destabilizes the protein structure, and hence leads to an increase in oligomer size. 

Similar to other small heat shock proteins (sHSPs), α-crystallin also contains a highly conserved sequence of 80 to 100 residues (residue 62-143 in αA-crystallin and 66-147 in αB-crystallin) called the α-crystallin domain. ^{40 41} Based on similarities with the structure of other HSPs, it is believed that the N-terminal region (residue 1-62 in αA-crystallin and 1-66 in αB-crystallin) of α-crystallin forms an independently folded domain, whereas the C-terminal (referred as the C-terminal extension; residues 143-173 in αA- and 147-175 in αB-crystallin) is flexible and unstructured. ⁴¹ Both deamidation at N₇₈ and N₁₄₆ are within the α-crystallin domain region (residue 66-147) of αB-crystallin. The α-crystallin domain is engaged in the subunit–subunit interactions, because recombinant αB-crystallin containing only the α domain region forms a dimer. ⁴² Both N₇₈ and N₁₄₆ are important for subunit interaction and chaperone activity. Between the two, the N₁₄₆ residue is relatively critical to maintenance of chaperone activity and for proper oligomer sizes of αB homopolymers. How deamidation of either site (N₇₈ and N₁₄₆) would affect interaction of αA- and αB-subunits remains to be determined. By and large, attempts to identify individual amino acids in subunit interactions and chaperone activity have been unsuccessful, because site-directed mutagenesis did not cause extensive perturbation in the crystallin structure. However, two disease-related point mutations of a highly conserved Arg at equivalent positions in αA (R116C) and αB (R120G) cause structural changes that lead to hereditary cataracts ^{43 44 .} Deletion of the last 17 amino acids from human αB-crystallin causes precipitation, with reduced chaperone activity, ⁴⁵ and a deletion of 25 residues from the C-terminus in Xenopus Hsp30c reduces its solubility and impairs chaperone activity. ⁴⁶  

Together, results of these studies have shown ^{28 43 45 46} that N- and C-terminal regions are essential for proper folding of α-crystallin, subunit interactions between αA- and αB-crystallins, and chaperone activity. However, the C-terminal regions seem to be needed to preserve the native structure of the molecule. ⁴⁶ It is presently unknown whether deamidation of N₇₈ and/or N₁₄₆ affects the role of N- and C-terminal extensions of α-crystallin. Further, as stated earlier, deamidation may serve as a signal for proteolysis. Whether this signal is used during age- and cataract-related truncations of αA- and αB-crystallins and other crystallins remains to be determined. We are presently attempting to find answers to these questions. 

 Supported by National Eye Institute Grant EY06400.

 Submitted for publication July 10, 2003; revised September 24, 2003; accepted October 1, 2003.

 Disclosure: R. Gupta, None; O.P. Srivastava, 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.

 

Table 1.

 

Oligonucleotide Primers Used in Mutagenesis of N₁₄₆, N₇₈, and N₇₈/N₁₄₆ Residues in Human Lens αB-Crystallin

Constructs	Primers
N146D
Forward	5′GGGGTCCTCACTGTGGACGGACCAAGGAAACAGG3′
Reverse	5′CCTGTTTCCTTGGTCCGTCCACAGTGAGGACCCC3′
N78D
Forward	5′GACAGGTTCTCTGTCGACCTGGATGTCAAGCAC3′
Reverse	5′GTGCTTCACATCCAGGTCGACAGAGAACCTGTC3′

Figure 1.

 

Isotopic distribution of the MALDI-TOF spectrum of tryptic fragments from WT αB-crystallin and deamidated mutants. (A) A tryptic fragment of WT αB-crystallin with a mass of 921.57 (residue 75-82). (B) A tryptic fragment of WT αB-crystallin with a mass of 2625 (residue 124-149). (C) The N78D mutant, showing a mass of 922.56 (residue75-82) with a gain of 1 mass unit due to deamidation and (D) the N146D mutant showing a mass of 2626.08 (residue 124-149), an increase of 1 mass unit after deamidation of residue N₁₄₆. The N78D/N146D mutant showed a mass per charge of 922.56 and 2626.08 as shown in (C) and (D) after deamidation at N₇₆ and N₁₄₆ (results not shown).

Figure 2.

 

SDS-PAGE analysis of purified WT αB-crystallin and the three deamidated mutants. Purification of the WT and the three mutant αB-crystallins was performed by a combination of methods (anion-exchange and hydrophobic interaction chromatography).

Figure 3.

 

Comparison of chaperone activities of the WT αB-crystallin and three deamidated mutants. The chaperone activity assays were performed with three separate substrates and three different temperatures. The micromolar concentrations of the WT species and the N78D, N146D, and N78D/N146D mutants of αB-crystallin were based on the aggregate molecular mass of individual protein, as determined by light-scattering. Protein aggregation in each assay was determined by monitoring light-scattering at 360 nm. Chaperone activity is represented as the percentage protection provided by the WT and mutant proteins. (A) Aggregation of insulin B chain by reduction with DTT in the presence of different chaperonin-to-insulin ratios at 25°C. The concentration of insulin was 1 μM. (B) Aggregation of CS in the presence of different chaperonin-to-CS ratios at 43°C. The concentration of CS was 1 μM. (C) Aggregation of human recombinant γD-crystallin in the presence of different chaperonin-to-γD crystallin ratios at 63°C. The concentration of recombinant γD crystallin was 1 μM.

Figure 4.

 

Fluorescence spectra of WT and three deamidated mutants of αB-crystallin after ANS binding. Fluorescence spectra of ANS bound to WT and the three deamidated mutants at (A) 37°C and (B) 43°C. The concentration of each αB-crystallin species was 0.15 mg/mL, and that of ANS was 12 μM. WT and mutated protein preparations were heated with ANS at 43°C for 10 minutes and fluorescence spectra determined after cooling them to room temperature. Arrows: WT αB-crystallin and the three deamidated mutants.

Figure 5.

 

Intrinsic Trp fluorescence spectra of WT and three deamidated mutants of αB-crystallin. The preparations were excited at 295 nm and fluorescence spectra were recorded between 300 and 400 nm of emission. The concentration of each crystallin species was 0.15 mg/mL. Arrows: WT αB-crystallin and the three deamidated mutants.

Figure 6.

 

Far-UV circular dichroism spectra of WT and three deamidated mutants of αB-crystallin. The spectra were recorded at the concentration of 1 mg/mL of each αB species and a cell path length of 0.1 cm. Arrows: WT αB-crystallin and the three deamidated mutants.

Figure 7.

 

SLS measurements of WT and three deamidated mutants of αB-crystallin. Elution profile and molecular mass of oligomers formed by WT and the deamidated mutants (arrows). The concentration of the αB-crystallin species was 1 mg/mL.

Table 2.

 

SLS Determination of Molecular Weights of Oligomers of WT αB-Crystallin and Three Deamidated Mutants (N78D, N146D, and N78D/N146D)

αB-Crystallin Species	Molecular Weight
WT	580,000
N78D	670,000
N146D	750,000
N78D/N146D	770,000

The authors thank Martha Robbins for editorial assistance and Donald Muccio, PhD, and Mike Jabalonski for technical assistance and recording the far-UV CD spectra in the Chemistry Department of the University of Alabama at Birmingham (UAB); and John Quinn for technical assistance in determining the molecular weight by SLS (Department of Microbiology, UAB).