Subunit Exchange of αA-Crystallin*

α-Crystallin, the major protein in the mammalian lens, is a molecular chaperone that can bind denaturing proteins and prevent their aggregation. Like other structurally related small heat shock proteins, each α-crystallin molecule is composed of an average of 40 subunits that can undergo extensive reorganization. In this study we used fluorescence resonance energy transfer to monitor the rapid exchange of recombinant α-crystallin subunits. We labeled αA-crystallin with stilbene iodoacetamide (4-acetamido-4′-((iodoacetyl)amino)stilbene-2,2′-disulfonic acid), which serves as an energy donor and with lucifer yellow iodoacetamide, which serves as an energy acceptor. Upon mixing the two populations of labeled αA-crystallin, we observed a reversible, time-dependent decrease in stilbene iodoacetamide emission intensity and a concomitant increase in lucifer yellow iodoacetamide fluorescence. This result is indicative of an exchange reaction that brings the fluorescent αA-crystallin subunits close to each other. We further showed that the exchange reaction is strongly dependent on temperature, with a rate constant of 0.075 min−1 at 37 °C and an activation energy of 60 kcal/mol. The subunit exchange is independent of pH and calcium concentration but decreases at low and high ionic strength, suggesting the involvement of both ionic and hydrophobic interactions. It is also markedly reduced by the binding of large denatured proteins. The degree of inhibition is directly proportional to the molecular mass and the amount of bound polypeptide, suggesting an interaction of several αA-crystallin subunits with multiple binding sites of the denaturing protein. Our findings reveal a dynamic organization of αA-crystallin subunits, which may be a key factor in preventing protein aggregation during denaturation.

␣-Crystallin, the major protein in the mammalian lens, is a molecular chaperone that can bind denaturing proteins and prevent their aggregation. Like other structurally related small heat shock proteins, each ␣-crystallin molecule is composed of an average of 40 subunits that can undergo extensive reorganization. In this study we used fluorescence resonance energy transfer to monitor the rapid exchange of recombinant ␣-crystallin subunits. We labeled ␣A-crystallin with stilbene iodoacetamide (4-acetamido-4-((iodoacetyl)amino)stilbene-2,2disulfonic acid), which serves as an energy donor and with lucifer yellow iodoacetamide, which serves as an energy acceptor. Upon mixing the two populations of labeled ␣A-crystallin, we observed a reversible, time-dependent decrease in stilbene iodoacetamide emission intensity and a concomitant increase in lucifer yellow iodoacetamide fluorescence. This result is indicative of an exchange reaction that brings the fluorescent ␣Acrystallin subunits close to each other. We further showed that the exchange reaction is strongly dependent on temperature, with a rate constant of 0.075 min ؊1 at 37°C and an activation energy of 60 kcal/mol. The subunit exchange is independent of pH and calcium concentration but decreases at low and high ionic strength, suggesting the involvement of both ionic and hydrophobic interactions. It is also markedly reduced by the binding of large denatured proteins. The degree of inhibition is directly proportional to the molecular mass and the amount of bound polypeptide, suggesting an interaction of several ␣A-crystallin subunits with multiple binding sites of the denaturing protein. Our findings reveal a dynamic organization of ␣A-crystallin subunits, which may be a key factor in preventing protein aggregation during denaturation.
␣-Crystallin, the major lens protein of the mammalian eye, is a member of the small heat shock protein family (1,2). Like other small heat shock proteins, ␣-crystallin is a high molecular mass complex consisting of a large number of subunits. The two polypeptides of ␣-crystallin found in the lens of the mammalian eye, ␣A and ␣B, are encoded by evolutionarily related genes and share more than 50% identity in amino acid sequence (3,4). For many years, ␣-crystallin was thought to be lens-specific. However, recent advances in detection methods have revealed much wider non-lenticular tissue distributions in heart, thymus, skin, lung, retina, and brain (5)(6)(7)(8).
In the past, ␣-crystallin was considered to play only a structural role in maintaining the transparency of the lens. Recent studies have demonstrated that ␣-crystallin can prevent the thermally induced aggregation of a diversity of denaturing proteins (9 -14). These findings suggest that ␣-crystallin possesses chaperone-like property that prevents aggregation of denatured lens proteins, thus preserving the transparency of the lens and reducing the probability of developing cataract (15). Its protective function is further supported by the detection of an elevated amount of ␣B-crystallin under heat and hypertonic stress (16 -18) and in a number of neurological degenerative diseases including Creutzfeldt-Jacob disease (19,20), diffuse Lewy body disease (21), Alzheimer's disease (22), and Alexander's disease (23). However, the exact physiological role of ␣-crystallin in these diseases is still unclear.
␣-Crystallin is normally isolated as an oligomeric complex with an average of 40 subunits and a molecular mass of 8 ϫ 10 5 Da (3). However, its size distribution can vary from 3 ϫ 10 5 to 1.5 ϫ 10 8 Da, depending on the age of the tissue from which it is isolated (24 -29), and the temperature, calcium concentration, pH, and ionic strength of the assay conditions (30 -33). Moreover, calf lens ␣-crystallin that was separated into five subpopulations with distinct molecular masses has been shown to rapidly return to its original distribution upon mixing (34). Exchange of subunits between native and phosphorylated forms of ␣-crystallin has also been detected by isoelectric focussing (35). These findings suggest that the subunits of ␣crystallin are capable of freely associating and dissociating to form large multimeric protein complexes.
In this study we developed a fluorescence resonance energy transfer method to monitor the exchange of recombinant ␣Acrystallin subunits. Using this technique, we have determined the effect of pH, Ca 2ϩ , and ionic strength on the rate of subunit exchange. We further found a strong dependence of subunit exchange on temperature, with an activation energy of 60 kcal/mol. Binding of large denatured proteins to ␣A-crystallin markedly reduced the exchange rate, indicating an association of the polypeptides with several ␣A-crystallin subunits. The multiple interactions may explain why the binding of denatured proteins to ␣A-crystallin is irreversible.

EXPERIMENTAL PROCEDURES
Materials-Lucifer yellow iodoacetamide (LYI) 1 and 4-acetamido-4Ј-((iodoacetyl)amino)-stilbene-2,2Ј-disulfonic acid (AIAS) were purchased from Molecular Probes, Eugene, OR. Ovotransferrin, ␣-lactalbumin, insulin, and melittin were obtained from Sigma. They were used in the experiments without further purification. Restriction enzymes and Taq polymerase were purchased from New England Biolabs and Promega, respectively. Escherichia coli strain BL21DE3, pT7 Blue T-cloning vector and the pET 20b ϩ expression vector were obtained from Novagen. Rat lens epithelial gt-11 cDNA library was a generous gift of Dr. S. Bhat, Jules Stein Eye Institute, UCLA.
Cloning of ␣A-Crystallin from Rat Lens Epithelium-cDNA of a␣-crystallin was obtained from the rat lens epithelial gt-11 cDNA library by polymerase chain reaction amplification using sense primer 5Ј-TCACCATCCAGCACCCTTG-3Ј and antisense primer 5Ј-TCAGGAC-GAGGGTGCCGAG-3Ј. The polymerase chain reaction reaction was carried out in a 25-l volume containing 10 mM Tris-HCl, pH 8.3, 50 mM KCl, 1.5 mM MgCl 2 , 50 M dNTP, 1.0 M primers, 1 l of cDNA library, and 2.5 units of Taq polymerase. Amplification was performed for 35 cycles with conditions for denaturation at 94°C for 1 min, annealing at 54°C for 2 min, and extension at 72°C for 3 min. The 525-base-long polymerase chain reaction product was gel-purified, ligated into pT7 blue vector, and subsequently subcloned into the pET 20b ϩ expression vector. This construct introduced a conservative substitution of Val-3 3 Phe near the amino terminus. Comparison of the DNA sequence of our construct with that of the published rat sequence (36) also indicated a single base change converting Ser-129 3 Cys. Both errors were subsequently corrected by mutagenesis.
Expression and Purification of Recombinant ␣A-Crystallin-BL21DE3 cells containing the pET 20b ϩ ␣A-crystallin plasmids were grown in 500 ml of LB broth to a cell density between 0.6 and 1.0 A at 600 nm and induced with 0.5 mM isopropyl-␤-D-thiogalactopyranoside for 3 h. Cells were harvested by centrifugation at 4,000 ϫ g for 10 min, resuspended in 20 ml of ice-cold buffer containing 50 mM Tris, pH 7.9, 0.1 M NaCl, 2 mM EDTA and lysed by sonication. The cell particulates were removed by centrifugation at 15,000 ϫ g for 30 min, followed by filtration through a 0.2 m filter. Polyethyleneimine was then added to the filtrate with rapid stirring to form a 0.12% solution. After incubation on ice for 2 min, the mixture was centrifuged at 15,000 ϫ g for 10 min to remove the precipitated DNA. The supernatant was adjusted with DTT to a final concentration of 10 mM and applied onto a Mono-Q column (Pharmacia Biotech Inc.) pre-equilibrated in 100 mM NaCl, 20 mM Tris-HCl, pH 8.5. Proteins were eluted using a linear gradient of 0.1-1 M NaCl in the same buffer. Fractions containing recombinant ␣A-crystallin were concentrated and applied to a Superose 6 gel filtration column (Pharmacia) equilibrated in 100 mM NaCl, 20 mM Tris, pH 7.9. Fractions containing purified ␣A-crystallin were pooled, concentrated to 20 mg/ml, and stored frozen at Ϫ20°C.
Labeling of Recombinant ␣A-Crystallin with Fluorescence Probes-Recombinant ␣A-crystallin (20 mg/ml) was diluted to 1 mg/ml with buffer containing 100 mM NaCl, 20 mM MOPS, pH 7.9. Solid AIAS was then added to a final concentration of 3.2 mM, and the reaction was allowed to proceed for 12 h at room temperature (22°C) in the dark. Unreacted AIAS was separated from the fluorescently labeled ␣A-crystallin on a G-25 Sephadex desalting column (Pharmacia) equilibrated with buffer A (100 mM NaCl, 2 mM DTT, 50 mM sodium phosphate, pH 7.5). ␣A-Crystallin was covalently labeled with LYI at a final concentration 8.4 mM under the same conditions, except that the reaction was extended for an additional 6 h at 37°C.
Measurements of the Rate of Subunit Exchange-Fluorescence energy transfer was employed to determine the rate of subunit exchange. The exchange reaction was initiated by mixing an equal volume of 0.4 mg/ml AIAS-labeled ␣A-crystallin and 0.4 mg/ml LYI-labeled ␣A-crystallin at 37°C in buffer A. At time 0, 10, 25, 60, 120, 180, and 240 min, 20 l of the reaction mixture was removed and diluted 100 ϫ with the same buffer. The emission spectrum of the sample excited at 335 nm was recorded using an Perkin-Elmer LS-5 spectrofluorometer, and the intensity at 415 nm was determined. The rate of subunit exchange was calculated from the equation is the fluorescence intensity at 415 nm and k is the rate constant of subunit exchange. The constants, C 1 and C 2 were determined using the conditions where C 1 ϩ C 2 ϭ 1 at time 0 and C 1 is fluorescence intensity at time ϱ. The rate constant was determined by nonlinear regression analysis of the data using the Biomedical Statistical Package program.
Determination of the Effect of pH, Ionic Strength, and Ca 2ϩ -The effect of pH on subunit exchange was determined using the same procedure, except that the measurements were carried out either in 50 mM sodium phosphate, 100 mM NaCl, pH 6.5, 7.5, or 8.0, or in 50 mM sodium borate, 100 mM NaCl, pH 9.2. To determine the effect of ionic strength or Ca 2ϩ , the ␣A-crystallin solution was exhaustively dialyzed against 10 mM MOPS, pH 7.5, and adjusted with 2 M NaCl or 1 M CaCl 2 solution to the ion concentration as indicated in the figure legends. The fluorescence intensity of the sample was determined immediately and at 15 min after mixing an equal volume of 0.4 mg/ml AIAS-labeled ␣A-crystallin and 0.4 mg/ml LYI-labeled ␣A-crystallin at 37°C. The rate of subunit exchange in these experiments is defined as the change in relative fluorescence intensity at 415 nm in 15 min.
Determination of the Effect of Bound Polypeptides-The effect of bound polypeptides on subunit exchange were determined with four different proteins that are known to bind tightly to ␣-crystallin under denaturing conditions. The binding was performed by incubating different concentrations of protein with either 0.4 mg/ml AIAS-labeled or LYI-labeled ␣A-crystallin. The denaturing conditions were as follows: melittin, 0.1 M NaCl, 50 mM sodium phosphate, pH 7.5, at room temperature for 30 min (37); insulin, 20 mM DTT, 50 mM sodium phosphate, pH 7.0, at room temperature for 60 min (36); ␣-lactalbumin, 2 mM EDTA, 50 mM DTT, 0.1 M NaCl, 50 mM sodium phosphate, pH 7.0, at 37°C for 90 min; ovotransferrin, 20 mM DTT, 0.1 M NaCl, 50 mM sodium phosphate, pH 7.9, at 42°C for 90 min. After binding, equal amounts of AIAS-labeled and LYI-labeled ␣A-crystallin containing the bound polypeptide were incubated at 37°C for 15 min, and the relative fluorescence intensity of the preparation was determined as described in the previous section.
Analytical Methods-Protein concentrations were determined by Coomassie Blue binding (38) using ␥-globulin as a standard. SDSpolyacrylamide gel electrophoresis of proteins was performed by the method of Laemmli (39). The concentrations of LYI and AIAS were determined from their absorption spectra using molar extinction coefficients of 13,000 cm Ϫ1 M Ϫ1 at 435 nm and 35,000 cm Ϫ1 M Ϫ1 at 335 nm, respectively.

RESULTS
Expression of Recombinant ␣A-Crystallin-␣A-Crystallin cDNA was constructed in vector pET 20b ϩ carrying a strong bacteriophage T7 promotor. The resulting expression construct was used to transform E. coli BL21DE3 cells containing a chromosomal copy of the inducible T7 RNA polymerase gene. The level of expression of ␣A-crystallin in this system was greater than 50% that of the total proteins. When the cell lysate was separated into soluble and particulate fractions by centrifugation, more than 90% of the recombinant ␣A-crystallin was found in the soluble fraction (data not shown). The high expression level of ␣A-crystallin in the soluble fraction allowed purification of the recombinant protein to greater than 95% purity by successive Mono-Q ion exchange chromatography and gel filtration chromatography.
Purified recombinant ␣A-crystallin and native lens ␣A-crystallin were equally effective in preventing the thermally-induced aggregation of proteins. Their conformations, as determined by circular dichroism, were identical (data not shown). They also interacted to form a multimeric subunit complex of 800 kDa, which is similar in size to native ␣-crystallin isolated from lens.
Labeling of ␣A-Crystallin with Fluorescent Probes-Recombinant ␣A-crystallin contains a single cysteine residue at position 131 that can be used to attach a sulfhydryl-specific fluorophore. Fig. 1 shows the absorption spectra of AIAS-labeled ␣A-crystallin (upper panel) and LYI-labeled ␣A-crystallin (lower panel) obtained by the method given under "Experimental Procedures." Calculations based on the molar extinction coefficent at 335 nm for AIAS and 435 nm for LYI revealed an average of 1 mol of fluorophore/mol of ␣A-crystallin subunit, suggesting the covalent attachment of the fluorophore to a single site at Cys-131.
The modification of Cys-131 did not appear to perturb the conformation or the interaction of the ␣A-crystallin subunits. Fig. 2 shows a comparison of the gel filtration profiles between unlabeled ␣A-crystallin and AIAS-labeled ␣A-crystallin. Their average molecular masses were both 800 kDa, and their size distribution ranged from 300 to 1,000 kDa. Similar size distribution was obtained with LYI-labeled ␣A-crystallin.
The emission spectra of AIAS-labeled ␣A-crystallin excited at 335 nm (upper panel), and LYI-labeled ␣A-crystallin excited at 435 nm (lower panel) are shown in Fig. 1. The emission maxima of AIAS-labeled ␣A-crystallin and LYI-labeled ␣Acrystallin were at 415 and 525 nm, respectively. The significant overlap of the emission spectrum of the AIAS fluorophore with the absorption band of the LYI fluorophore indicates that they are an excellent donor-acceptor pair for fluorescence resonance energy transfer.

Determination of Subunit Exchange by Fluorescence Resonance Energy
Transfer-Both AIAS-labeled and LYI-labeled ␣A-crystallin were very stable at 37°C, with no significant change in fluorescence intensity over a period of 12 h. If the oligomeric complex of ␣A-crystallin is static and subunit ex-change does not occur after mixing the two populations of labeled ␣A-crystallin, the fluorescence intensity will remain the same, since the donor and the acceptor are far apart. However, the fluorescence intensity was markedly altered upon mixing of the two populations of labeled ␣A-crystallin (Fig. 3). The time-dependent decrease in AIAS emission intensity at 415 nm and a concomitant increase in LYI fluorescence at 525 nm were indicative of energy transfer due to the close proximity of the two fluorophores. The quenching of the fluorescence was completed in 4 h at 37°C (Fig. 4), resulting in approximately 40% decrease of the original AIAS fluorescence intensity.
The rate of subunit exchange can be obtained by measuring either the decrease in donor fluorescence or the increase in acceptor fluorescence. The upper panel of Fig. 4 shows a plot of the emission intensity of AIAS at 415 nm as a function of time after the mixing of the two populations of labeled ␣A-crystallin. An exchange rate constant of 0.075 min Ϫ1 was obtained by fitting the data to the exponential function F(t) ϭ C 1 ϩ C 2 e Ϫkt . The same exchange rate was determined by measuring the increase in LYI fluorescence intensity at 545 nm (lower panel). Since both measurements gave essentially the same rate constant, all subsequent measurements were obtained by monitoring only the quenching of the AIAS fluorescence.
Reversibility of Subunit Exchange- Fig. 5 shows that the AIAS fluorescence at 415 nm rapidly recovered upon the addition of unlabeled ␣A-crystallin to a pre-mixed population of AIAS-labeled and LYI-labeled ␣A-crystallin at 37°C. This result demonstrates that the subunit exchange reaction is reversible.
Effects of pH-The effect of pH on the subunit exchange of ␣A-crystallin at 37°C is shown in Fig. 6. For this experiment, the rate of subunit exchange was determined by the change in relative fluorescence intensity in 15 min after mixing an equal amount of AIAS-labeled ␣A-crystallin and LYI-labeled ␣A-crystallin. Except for a small decrease in the relative fluorescence intensity at pH 6.5, the subunit exchange rate at pH 7.5, 8.0, and 9.2 were the same. Effects of Sodium and Calcium Ions-Calcium ions have been shown previously to change the subunit organization of ␣-crystallin (33). Fig. 7 shows that calcium chloride concentration in the range of 0 -50 mM had no effect on subunit exchange.
The effect of NaCl on the subunit exchange of ␣A-crystallin is more complex (Fig. 8). Increasing the NaCl concentration from 0 to 0.1 M resulted in an increase in subunit exchange, as indicated by a decrease in relative fluorescence intensity at 15 min. However, at 1.0 M NaCl, a small decrease of the exchange rate was observed. These results suggest that the multimeric subunit organization of ␣A-crystallin may be stabilized by both ionic interaction and hydrophobic interactions.
Effect of Temperature-The rate of subunit exchange was highly temperature-dependent. The AIAS-labeled and LYI-labeled ␣A-crystallin exchanged at a rate 4.2-fold higher at 42°C than at 37°C (Fig. 9, upper panel), and the exchange took only 21 min to complete. In contrast, when the temperature was reduced to 3°C, subunit exchange was not detectable over a period of 6 h. Fig. 9 (lower panel) shows the Arrhenius plot of the subunit exchange reaction. The activation energy for subunit exchange, as determined from the slope of the plot of ln(k) versus 1/T in degrees K was 60 kcal/mol.
Effect of Bound Denatured Proteins-␣A-Crystallin has been shown to bind irreversibly to a number of denatured proteins and prevent their aggregation (9 -14). What is the effect of bound polypeptide on the rate of subunit exchange? To answer this important question, AIAS-labeled and LYI-labeled ␣Acrystallin were incubated separately with different amounts of proteins under denaturing conditions. After binding, the rate of  subunit exchange was determined, and the changes in exchange rate were then plotted as a function of the protein concentration during denaturation. Fig. 10 shows a comparison of the exchange rate of ␣A-crystallin containing either bound melittin, insulin B chain, ␣-lactalbumin, or ovotransferrin. Binding of short polypeptides such as melittin (2.6 kDa) and insulin B chain (3.0 kDa) to ␣A-crystallin either did not alter the exchange rate or had only a small effect at high concentrations. In contrast, binding of a large polypeptide such as ovotransferrin (40 kDa) markedly reduced the rate of subunit exchange. The degree of inhibition is proportional to the amount of bound ovotransferrin, with a 35% decrease in exchange rate at a 2:1 molar ratio of ␣A-crystallin subunit to ovotransferrin. There is also a close correlation between the size of the bound polypeptide and the inhibition of subunit exchange. Bound ␣-lactalbumin (14 kDa) was found to be more effective than insulin but less effective than ovotransferrin in inhibiting subunit exchange. These results suggest the association of larger polypeptides with several ␣A-crystallin subunits, effectively cross-linking them together and preventing them from exchanging.

DISCUSSION
In this study we have used fluorescence resonance energy transfer (40,41) to monitor the rapid exchange of ␣A-crystallin subunits. We demonstrated that the exchange can be fit to a simple exponential function with a rate constant of 0.075 min Ϫ1 at 37°C and reached a complete equilibrium within 4 h (Fig. 4). Our results are similar but not identical to those of van den Oetelaar et al. (35), who measured the mixing of isolated bovine ␣Aand ␣B-crystallins by isoelectric focussing (35). In this earlier study, a prolonged exchange reaction between ␣Aand ␣B-crystallins that did not completely terminate until after 24 h was observed. This difference can be accounted for by a number of contributing factors. First, the exchange between ␣Aand ␣B-crystallin subunits may be intrinsically slower, a hypothesis that we are currently investigating. Second, the ␣-crystallin preparations may be crucial to the subunit exchange measurement. The native bovine ␣A-crystallin used in the earlier study of van den Oetelaar et al. our AIAS-labeled and LYI-labeled ␣A-crystallin by the modification of the cysteine residue. Finally, the assay conditions may also have an effect on the kinetics of the exchange reaction. For example, we have found that the rate of subunit exchange is noticeably slower (Fig. 8) under the lower ionic strength conditions used in earlier experiments (35).
Of all the biochemical parameters we have examined, temperature has the most pronounced effect on subunit exchange of ␣A-crystallin. The exchange rate increased markedly upon increasing the temperature from 37 to 42°C (Fig. 9). This result is important in light of the potential roles of ␣-crystallin in transcriptional regulation. Although ␣-crystallin does not have a definitive nuclear localization sequence, it has been shown to translocate from the cytoplasm to the nucleus of NIH 3T3 cells under heat shock conditions (16). ␣-Crystallin has also been found to bind specifically downstream of the transcription initiation site on the ␥-crystallin promoter (42). How does ␣-crystallin, which has an average molecular mass of 800 kDa, cross the nuclear membrane with an exclusion limit of approximately 70 kDa (43,44)? We propose that the dissociation of ␣-crystallin into smaller subunits at high temperatures may explain its entry into the nucleus. Under normal physiological condition, subunit dissociation is largely prevented by the high activation energy barrier of 60 kcal/mol for the subunit exchange reaction, which would explain why nuclear localization is predominantly observed only at higher temperatures (16).
The interaction of ␣A-crystallin subunits most likely involves both ionic and hydrophobic interactions, since we have observed a decrease in the exchange rate under either very low salt or very high salt conditions. The effect of ionic strength on subunit exchange was not detected previously by isoelectric focussing (35). The negative result can be readily explained by the fact that these earlier experiments were all performed at low salt conditions to produce sharply focused protein bands. As a result, only a relatively narrow range of salt concentrations has been tested, which falls outside the region where the ionic interaction becomes significant.
Surprisingly, although changes in pH have been implicated to have a considerable effect on the subunit organization of ␣-crystallin (30, 31), we have found that the exchange rate is constant at pH values ranging from 7.5 to 9.2 (Fig. 7). Our result is in agreement with van den Oetelaar et al. (35), who also showed that subunit exchange is independent of pH.
Ca 2ϩ is another biochemical parameter that we thought would have an influence on subunit exchange, since previous studies have reported the detection of large light scattering aggregates of ␣-crystallin when bovine lens is incubated in solution containing 4 -8 mM Ca 2ϩ (33,45). Instead, we found that the exchange rate is independent of Ca 2ϩ at concentrations as high as 50 mM (Fig. 7). It is not clear why ␣-crystallin in the lens is more susceptible to Ca 2ϩ -induced aggregation. One possible explanation is a change in subunit-subunit interaction due to post-translational modifications as the lens ages (46,47), an effect that is absent with recombinant ␣A-crystallin.
Which region of the ␣A-crystallin molecule is involved in subunit exchange? Although several mutational analyses of ␣-crystallin have been reported, very little is known about the contact sites between its subunits (48,49). Based on the high activation energy of 60 kcal/mol relative to other exchange reactions (50), we speculate that the subunit-subunit interactions may involve multiple binding sites. Recently, a stretch of 35 amino acid residues of Hsp42 has been implicated in subunit-subunit interaction (51). Comparison of the amino acid sequence of ␣A-crystallin to that of Hsp42 suggests residues 112-147 may be involved in the same function. Site-directed mutagenesis of ␣A-crystallin is currently under way to answer this question.
␣-Crystallin has been shown to possess chaperone-like property that prevents protein aggregation during denaturation (9 -14). Our study indicates that subunit exchange is not significantly affected by the binding of small polypeptides such as melittin or insulin B chain. In contrast, when ␣-lactalbumin or ovotransferrin bind to ␣A-crystallin, the subunit exchange rate is markedly reduced. This result implicates an association of larger polypeptides with several ␣A-crystallin subunits, effectively cross-linking them together and preventing them from dissociating. The multiple interactions may explain why the binding of denatured proteins to ␣A-crystallin is so strong and irreversible (9,52).
Although the exact mechanism of chaperone-like activity is still not clear, it is tempting to speculate that ␣-crystallin most likely recognizes certain structures that are transiently exposed during unfolding of the protein. The domains for binding denatured protein and for subunit-subunit interaction are distinct, since ␣-crystallin containing bound proteins retains its multimeric subunit organization (37,52). Moreover, ␣-crystallin is known to bind insulin B chain at room temperature (37), under which subunit exchange is largely inhibited (Fig. 9). If protein binding and subunit-subunit contact sites are located in a different part of the ␣-crystallin molecule, is there a relationship between subunit exchange and chaperone function? It is tempting to speculate that the rearrangement of ␣-crystallin subunits is essential for covering the unfolded polypeptides, thus shielding them from aggregation. This hypothesis would explain why the subunit exchange reaction (Fig. 9) and chaperone-like activity are markedly enhanced at high temperature (53).
Small heat shock proteins have been shown to protect cells from stress (54 -57). Their ubiquitous tissue distribution, overexpression in a number of pathological states, and stress-induced cellular redistribution argue for their importance in safeguarding many important cellular processes. A hallmark of small heat shock proteins like ␣-crystallin is the large multimeric subunit organization (58,59), which we have shown here to undergo continuous rearrangement through the exchange of subunits. The challenge in the coming years will be to explain why the oligomeric structure is important for the function of small heat shock proteins.