Skip to main content
NCBI home page
Protein Science : A Publication of the Protein Society logoLink to Protein Science : A Publication of the Protein Society
. 2004 Aug;13(8):2223–2235. doi: 10.1110/ps.04627004

Probing folding and fluorescence quenching in human γD crystallin Greek key domains using triple tryptophan mutant proteins

Melissa S Kosinski-Collins 1, Shannon L Flaugh 1, Jonathan King 1
PMCID: PMC2279819  PMID: 15273315

Abstract

Human γD crystallin (HγD-Crys), a major component of the human eye lens, is a 173-residue, primarily β-sheet protein, associated with juvenile and mature-onset cataracts. HγD-Crys has four tryptophans, with two in each of the homologous Greek key domains, which are conserved throughout the γ-crystallin family. HγD-Crys exhibits native-state fluorescence quenching, despite the absence of ligands or cofactors. The tryptophan absorption and fluorescence quenching may influence the lens response to ultraviolet light or the protection of the retina from ambient ultraviolet damage. To provide fluorescence reporters for each quadrant of the protein, triple mutants, each containing three tryptophan-to-phenylalanine substitutions and one native tryptophan, have been constructed and expressed. Trp 42-only and Trp 130-only exhibited fluorescence quenching between the native and denatured states typical of globular proteins, whereas Trp 68-only and Trp 156-only retained the anomalous quenching pattern of wild-type HγD-Crys. The three-dimensional structure of HγD-Crys shows Tyr/Tyr/His aromatic cages surrounding Trp 68 and Trp 156 that may be the source of the native-state quenching. During equilibrium refolding/unfolding at 37°C, the tryptophan fluorescence signals indicated that domain I (W42-only and W68-only) unfolded at lower concentrations of GdnHCl than domain II (W130-only and W156-only). Kinetic analysis of both the unfolding and refolding of the triple-mutant tryptophan proteins identified an intermediate along the HγD-Crys folding pathway with domain I unfolded and domain II intact. This species is a candidate for the partially folded intermediate in the in vitro aggregation pathway of HγD-Crys.

Keywords: Human γD Crystallin, tryptophan, fluorescence, protein folding, folding intermediates, fluorescence quenching

Human mature-onset cataracts affect nearly 15% of the US population over 40 yr of age, and are the leading cause of blindness worldwide (National Eye Institute 2002). Pathological studies of cataractous lenses have revealed that cataracts are composed of protein aggregates that precipitate or polymerize in lens cells of the eye (Oyster 1999).

Human γD crystallin (HγD-Crys) is a protein synthesized during embryonic development that must remain soluble in the anucleated cells of the adult human eye lens for proper vision. Covalently modified HγD-Crys has been recovered in protein aggregates removed from aged, cloudy lenses (Hanson et al. 2000). HγD-Crys has 173 amino acids and shows high sequence and structural similarity to other γ-crystallins (Basak et al. 2003). Mutations in the gene encoding HγD-Crys have been found in families exhibiting juvenile-onset cataracts, further implicating HγD-Crys in cataractogenesis (Heon et al. 1999; Pande et al. 2001; Nadrut et al. 2003).

HγD-Crys is composed of antiparallel β-sheets arranged in four Greek-key motifs separated into two domains (Fig. 1 ▶). The two domains show high levels of structural and sequence conservation and appear to be the result of gene duplication during evolution (Wistow et al. 1983). Like most soluble γ-crystallins, HγD-Crys is monomeric in solution (Jaenicke 1999). HγD-Crys has four tryptophans that have been used to probe unfolding and refolding progression with fluorescence spectroscopy (Kosinski-Collins and King 2003).

Figure 1.

Figure 1.

Open in a new tab

Ribbon structure of wild-type human γD crystallin showing the location of the four native tryptophans at positions 42, 68, 130, and 156 (Basak et al. 2003).

The fluorescence signal of HγD-Crys increases when the protein is denatured in high concentrations of guanidine hydrochloride (GdnHCl). This indicates that the tryptophan fluorescence is quenched in the native state. Although this phenomenon has been observed in other proteins without metal ligands or cofactors (Eyles and Gierasch 2000; Lakowicz 1999; He et al. 2001), it is uncommon. Ultraviolet light has been proposed as one of the etiological agents of cataract formation (McCarty and Taylor 2002; Sasaki et al. 2002). Though the cornea absorbs a large fraction of the incident UV, the extraordinarily long lifetime of the lens crystallins raises the possibility of low-dose, long-term effects. The presence of tryptophan in the lens proteins may function in protecting the retina from ultraviolet light damage (Kurzel et al. 1973). In this case, fluorescence quenching may protect the lens proteins from ultraviolet light absorption. We have investigated which residues are involved in the quenching reaction to provide possible insight into the ability of the lens proteins to maintain stability and transparency over the human lifetime.

Human γD crystallin could be refolded to its native state at 37°C after dilution out of denaturant (Kosinski-Collins and King 2003). An in vitro aggregation pathway of HγD-Crys that competed with productive refolding was identified and may be related to the mechanism of its involvement in mature-onset cataracts (Kosinski-Collins and King 2003). The aggregation pathway was studied using atomic force microscopy and consisted of ordered, distinct intermediates. Using fluorescence spectroscopy, a partially folded hydrophobically collapsed intermediate was also identified in the productive refolding pathway. Within the lens, partially unfolded crystallin intermediates are likely to be recognized by α-crystallin (Bron et al. 2000; Cobb and Petrash 2002; Fu and Liang 2003).

Given the two-domain structure of HγD-Crys, it seemed possible that the intermediate identified in the aforementioned kinetic experiments had one intact domain and one partially unfolded domain. Previous studies of a bovine homolog of HγD-Crys, γB crystallin, showed that the protein could be denatured by urea at pH 2.0, but not at pH 7.0, and could be refolded over all reported concentrations of urea. The protein exhibited a three-stage transition in these equilibrium studies, representing sequential denaturation of the C-terminal and N-terminal domains at pH 2.0 (Rudolph et al. 1990; Mayr et al. 1997; Jaenicke 1999). Conversely, the closely homologous human γS crystallin, a protein also containing two domains with four Greek keys, unfolded and refolded without evidence of separate domain transitions (Wenk et al. 2000). HγD-Crys may possess differential domain stability that was not detected in the apparent two-state unfolding transition observed during equilibrium unfolding in GdnHCl. A partially folded intermediate with only one domain structured might be involved in the aggregation pathway of HγD-Crys. Such species may be involved in the protein deposition pathology of cataract.

We constructed triple-mutant tryptophan constructs, each containing only one of the four native tryptophans of HγD-Crys to determine the origin of the anomalous quenching signal and to provide reporters of conformation for different regions of the protein. This has made it possible to assess the thermodynamic stability and kinetic properties of the two domains.

Results

Purification of HγD-Crys

The cloned HγD-Crys gene (Pande et al. 2001) was excised from a pET3a plasmid and ligated into a pQE.1 plasmid as described in Materials and Methods. This plasmid added an N-terminal MKHHHHHHQ peptide to HγD-Crys. The addition of this peptide did not affect expression of HγD-Crys or the ability of the protein to fold into a native-like state during purification. In addition, we were not able to detect differences in the fluorescence or circular dichroism spectra (CD) of the His-tagged protein. The thermodynamic and kinetic unfolding and refolding properties of the His-tagged species were identical to the wild-type protein (results not shown).

Four triple-mutant proteins were constructed using site-directed mutagenesis. Each contained three tryptophan-to-phenylalanine substitutions retaining one native tryptophan (W42-only, W68-only, W130-only, and W156-only). All mutant plasmids were transformed into Escherichia coli M15(pREP4) cells and the proteins expressed during incubation at 37°C. Wild-type and all four triple-mutant proteins of HγD-Crys accumulated primarily in the soluble fraction of the cell lysates (>60%). These proteins were purified using the protocol developed for wild-type His-tagged HγD-Crys. The four mutant proteins behaved similarly to wild-type during the purification procedure.

Structure assignment

CD, native gel electrophoresis, and ultraviolet light absorbance were performed to assess the overall conformations of the mutant proteins.

Native gel electrophoresis of wild-type and the triple-mutant tryptophan constructs confirmed that all proteins retained similar native conformations. His-tagged wild-type HγD-Crys shows three distinct bands when separated on a native polyacrylamide gel. All mutant constructs exhibited similar native bands running with analogous mobility as wild type (results not shown).

Far-UV CD of the native proteins showed a primarily β-sheet structure for wild-type HγD-Crys and all four triple-mutant tryptophan constructs at 37°C (pH 7.0) (Fig. 2 ▶). All five proteins had a characteristic β-sheet minimum at 218 nm, although the amount of β-structure appeared to vary between the different constructs. In addition, none of the mutant proteins exhibited the inflection shown by wild-type HγD-Crys at 208 nm (Kosinski-Collins and King 2003), indicating that some of the secondary structure of wild type was diminished in the mutant proteins. The tryp-tophan-to-phenylalanine mutations may have slightly disrupted local β-sheet structures in the altered CD signal at 218 nm while maintaining overall conformations similar to wild type, as observed by native gels.

Figure 2.

Figure 2.

Open in a new tab

Far-UV CD of wild-type His-tagged HγD-Crys (diamonds), W42-only (solid squares), W68-only (open squares), W130-only (solid circles), and W156-only (open circles). Samples were prepared at a 300 μg/mL protein concentration and equilibrated in S buffer at 37°C.

The W68-only construct displayed the characteristic inflection of wild type observed in far-UV CD at 235 nm. W156-only showed a slight arc around this wavelength that was not as pronounced as W68-only. This may have been due to the overall decreased signal of W156-only. Therefore, it seems that the inflection in the wild-type spectrum at 235 nm reports environments around Trp 156 and Trp 68.

Ultraviolet light-absorbance scans in the peptide backbone region of native wild-type, W42-only, W68-only, W130-only, and W156-only HγD-Crys showed similar spectra between 190 and 240 nm at 37°C (pH 7.0) (results not shown). Ultraviolet light spectra in the region of tyro-sine and tryptophan absorption were also obtained from native and denatured proteins at 37°C (pH 7.0). All proteins had absorbance spectra typical of polypeptides containing high numbers of tryptophan and tyrosine residues, with a maximum at 276 nm. Absorption spectra of denatured states of the triple-mutant tryptophan mutants overlaid uniformly, indicating that all tryptophans were in similar denatured environments (Fig. 3A ▶). Ultraviolet light absorbance spectra of native triple-mutant tryptophan proteins exhibited the same overall shape and character, but had slightly different overall absorbance intensities (Fig. 3B ▶). This indicated that Trp 42, Trp 68, Trp 130, and Trp 156 were found in slightly different tertiary environments.

Figure 3.

Figure 3.

Open in a new tab

Aromatic residue ultraviolet light absorbance spectra of His-tagged HγD-Crys at 100 μg/mL protein in S buffer at 37°C. Wild-type (diamonds) and a representative triple mutant tryptophan protein, W42-only (solid squares) are shown denatured in 5.5 M GdnHCl (A). W68-only, W130-only, and W156-only exhibit denatured spectra indistinguishable from W42-only. Native protein aromatic absorbance is shown of wild-type His-tagged HγD-Crys (diamonds), W42-only (solid squares), W68-only (open squares), W130-only (solid circles), and W156-only (open circles). (B). Thirty percent of the data points are shown.

Fluorescence spectra

The overall character of the fluorescence emission spectra of wild-type, W42-only, W68-only, W130-only, and W156-only HγD-Crys were assessed by exciting the protein at 295 nm and observing fluorescence emission intensities from 310 to 420 nm in either S buffer or S buffer and 5.5 M GdnHCl (Fig. 4 ▶; Table 1). S buffer contained 10 mM NaPO4, 5 mM DTT, and 1 mM EDTA and was prepared at pH 7.0. Wild-type, His-tagged HγD-Crys had a native fluorescence emission maximum of 326 nm, and was quenched in the native state. The denatured protein exhibited a fluorescence emission maximum of 350 nm. This is similar to data previously reported for the non-His-tagged construct (Kosinski-Collins and King 2003; Fig. 4A ▶). Similarly, all of the triple-mutant tryptophan constructs had denatured fluorescence emission maxima of 350 nm. The native fluorescence emission maxima were 327 nm for W42-only, 329 nm for W68-only, 318 nm for W130-only, and 327 nm for W156-only (Table 1). W68-only and W156-only retained the native-state quenching observed in wild type (Fig. 4C,E ▶), whereas W42-only and W130-only were quenched in the denatured state (Fig. 4B,D ▶). W130-only is significantly more fluorescent in its native state than any of the other triple-mutant constructs.

Figure 4.

Figure 4.

Open in a new tab

Fluorescence emission of native (solid circles) and denatured (open circles) HγD-Crys. Protein was excited at 295 nm, and emission spectra were collected from samples of 10 μg/mL protein in S buffer or S buffer and 5.5 M GdnHCl at 37°C. Fluorescence spectra of wild-type His-tagged HγD-Crys (A), W42-only (B), W68-only (C), W130-only (D), and W156-only (E) are shown.

Table 1.

Equilibrium unfolding and refolding constants

Wild type W42-only W68-only W130-only W156-only
Native Fluorescence Emission Max 326 nm 327 nm 329 nm 318 nm 327 nm
Denatured Fluorescence Emission Max 350 nm 350 nm 350 nm 350 nm 350 nm
Exhibit Native State Quenching? Yes No Yes No Yes
Exhibit Aggregation at Low [GdnHCl]? Yes Yes Yes Yes Yes
[GdnHCl] at ½ Denaturation 2.8 M 1.3 M 1.3 M 2.0 M 2.0 M
[GdnHCl] at ½ Renaturation 2.1 M 1.3 M 1.3 M 2.0 M 1.7 M
Open in a new tab

Equilibrium refolding and unfolding

Equilibrium unfolding and refolding experiments were performed on the triple-mutant tryptophan constructs of HγD-Crys to determine the stability of the individual domains. Fluorescence spectra were collected as a function of Gd-nHCl concentration at 37°C as previously described (Kosinski-Collins and King 2003). All four proteins showed a midpoint for denaturant-induced unfolding in the range of from 1 to 2 M GdnHCl. All four proteins also exhibited reversible refolding above 1.0 M GdnHCl. However, upon dilution to lower concentrations of denaturant, the proteins exhibited a polymerization behavior similar to that described for wild-type HγD-Crys (Kosinski-Collins and King 2003). Solution turbidity measurements of refolded samples confirmed that the apparent increase in fraction-unfolded values at low GdnHCl concentrations was caused by signal obstruction due to aggregate formation (results not shown). The equilibrium unfolding and refolding results are summarized in Table 1.

The equilibrium unfolding and refolding data were similar for mutant proteins with tryptophans located in the same domain (Fig. 5 ▶). The midpoint of the denaturation transitions was 1.3 M for the two constructs that retain a native tryptophan in domain I (W42-only and W68-only) and 2.0 M for those in domain II (W130-only and W156-only). Only W156-only exhibited the slight hysteresis between unfolding and refolding observed for wild type, having a midpoint of renaturation of 1.7 M. The equilibrium unfolding and refolding transition of W42-only, W68-only, and W130-only did not exhibit hysteresis. Further studies of residues in domain II near Trp 156 may elucidate the discrepancies in the unfolding and refolding pathway that are causing the observed hysteresis.

Figure 5.

Figure 5.

Open in a new tab

Equilibrium unfolding and refolding of His-tagged HγD-Crys in GdnHCl. Tryptophan fluorescence was monitored during unfolding and refolding, and all samples were equilibrated in S buffer at a protein concentration of 10 μg/mL. A representative set of fraction unfolded data is shown for wild-type (A), W42-only (B), W68-only (C), W130-only (D), and W156-only (E). All protein data were analyzed using the ratio of fluorescence emission intensities at 360 nm over 320 nm. Fraction unfolded values were calculated from raw fluorescence intensity ratio measurements using the method described by Pace et al. (1989). Unfolding (open circles) and refolding (solid circles) transitions are presented for each protein at 37°C.

On the basis of resistance of W42-only and W68-only to solvent denaturation, domain I was less stable than domain II. Domain I exhibited an unfolding and refolding transition midpoint at 1.3 M GdnHCl, whereas domain II had a midpoint of 2.0 M GdnHCl. This indicated that wild-type equilibrium unfolding and refolding probably contained a partially denatured intermediate not readily visible in the fluorescence spectra of wild-type chains. However, it is important to note that if the tryptophan residues are vital to stability of HγD-Crys, triple substitutions would likely affect the overall thermodynamic parameters of the molecule. The altered transition midpoints of the mutant proteins may simply reflect a global destabilization of the molecule due to these changes.

Unfolding kinetics

The unfolding kinetics of HγD-Crys were studied by dilution of native protein into 5.5 M GdnHCl and S buffer at 37°C (pH 7.0), and monitoring the changes in the fluorescence emission (Fig. 6 ▶; Table 2). A syringe injection port that exhibited a dead-time of ~1 sec was used as the mechanism of dilution. Figure 6A ▶ shows the change in raw fluorescence signal for the four mutant proteins. Because the quenching characteristics vary for each of the mutants, fluorescence emission of W42-only and W130-only decreased upon denaturation, whereas emission of W68-only and W156-only increased upon unfolding. In Figure 6B ▶, the data have been normalized between the native and denatured states for ease of visual comparison. As shown in Figure 6A ▶, the major changes in fluorescence took place in the first 30 sec of the reaction. Millisecond time-scale intermedi-ate(s) may have formed within the dead-time of these experiments, and are not directly addressed here.

Figure 6.

Figure 6.

Open in a new tab

Unfolding kinetics of His-tagged HγD-Crys monitoring intrinsic tryptophan fluorescence with excitation at 295 nm. Emissions were monitored at 350 nm for wild type, W68-only, and W156-only, whereas emission at 320 nm was used for W42-only and W130-only. HγD-Crys was denatured by rapid dilution into 5.5 M GdnHCl at 37°C in S buffer to a final protein concentration of 10 μg/mL. A representative protein-unfolding time course is shown for wild-type His-tagged HγD-Crys (diamonds), W42-only (solid squares), W68-only (open squares), W130-only (solid circles), and W156-only (open circles). Data are shown over the first 150 sec of unfolding, although the data were collected over the entire unfolding process (~1 h). Raw fluorescence emission values (A) and data normalized between native and denatured emission intensities (B) are shown.

Table 2.

Unfolding kinetic rate constants

k1 t1/2 k2 t1/2 k3 t1/2
Wild type 0.2 3.5 sec 0.038 18 sec 0.0057 120 sec
W42-only 0.1 7.9 sec 0.039 18 sec NA NA
W68-only 0.2 3.5 sec 0.015 46 sec NA NA
W130-only NA NA 0.060 12 sec 0.0045 150 sec
W156-only NA NA 0.046 15 sec 0.019 36 sec
Open in a new tab

The kinetic data suggest the presence of partially unfolded intermediates in the transition between the native and unfolded states.

Wild-type His-tagged HγD-Crys was best fit with a four-state unfolding pathway (Fersht 1999). The protein started as a native species, and then formed two sequential intermediates before becoming completely unfolded. In this model, an early intermediate (Iu1) was populated within a t1/2 of 1.0 sec (Fig. 6 ▶). A second partially unfolded intermediate (Iu2) followed, forming with a t1/2 of 55 sec. This intermediate was not as quenched as native, indicating the polar–tryptophan interaction had been disrupted. A final unfolding transition (Iu2 →denatured) occurred with a t1/2 of 120 sec.

graphic file with name M1.gif (1)

The unfolding process for the domain I tryptophan constructs, W42-only and W68-only, exhibited two transitions corresponding to the early unfolding steps of wild type. These two proteins exhibited changes in fluorescence signal that best fit to a three-state unfolding model. Both proteins unfolded to the intermediate conformation with an initial t1/2 of 1.4 sec. W42-only had a subsequent transition from intermediate to denatured with a t1/2 of 18 sec and W68-only had a secondary t1/2 of 46 sec. These values represent rate constants that are similar to the k1 and k2 rate constants observed in wild type (Table 2).

By comparison, the changes in the fluorescence signals upon unfolding for the domain II tryptophan constructs, W130-only and W156-only, proceeded more slowly than the changes in emission for domain I. The unfolding curves for tryptophans from domain II were best fit to a three-state unfolding process involving a partially unfolded intermediate. Neither mutant had a k1 rate constant that was as large as that observed for wild type. Instead, the domain II mutants showed an unfolding process with rate constants that were similar to the k2 and k3 values of wild type. W130-only had t1/2 values of 12 and 150 sec for its two unfolding transitions, and W156-only had comparable values of 15 and 36 sec.

When comparing the tryptophan residues in homologous positions in the two domains, the unfolding curve trends were similar. The fluorescence signal observed for Trp 42 was similar to that of Trp 130, in that both had two unfolding transitions and populated an partially unfolded intermediate. The overall trend of the unfolding curve was similar for Trp 68 and Trp 156 as well, and both had a hyperfluo-rescent folding intermediate. This intermediate was likely a result of rapid relaxation of the tertiary structure surrounding Trp 68 and Trp 156, resulting in a partial release of the tryptophan–quencher interaction.

In addition, examination of tryptophans in homologous domain positions showed that domain I unfolded before domain II. Trp 42 had a more rapid change of global environment than Trp 130, whereas the fluorescence emission of Trp 68 increased its fluorescent signal more rapidly than Trp 156. Wild type exhibited both of these transitions and probably underwent sequential unfolding, in which domain I unfolded before domain II and an intermediate was populated with domain II folded, but domain I denatured.

Refolding kinetics

To study the kinetic rates and intermediates formed during refolding of HγD-Crys, fluorescence emission was monitored during refolding of denatured HγD-Crys at 37°C to a final denaturant concentration of 1.0 M GdnHCl in S buffer (Fig. 7 ▶; Table 3). Solution turbidity scans affirmed that no aggregate was formed under these conditions for any of the proteins described.

Figure 7.

Figure 7.

Open in a new tab

Refolding kinetics of His-tagged HγD-Crys monitoring intrinsic tryptophan fluorescence with excitation at 295 nm. Emissions were measured at 350 nm for wild type, W68-only, and W156-only, whereas emission wavelengths of 320 nm were used for W42-only and W130-only. HγD-Crys was denatured in 5.5 M GdnHCl at 37°C in S buffer for 3 h. HγD-Crys was refolded by rapid dilution with S buffer to a final GdnHCl concentration of 1.0 M and a final protein concentration of 10 μg/mL for wild-type His-tagged HγD-Crys (diamonds), W42-only (solid squares), W68-only (open squares), W130-only (solid circles), and W156-only (open circles). A representative scan of raw fluorescence signals (A) and signals normalized between native and denatured fluorescence values (B) are shown. Data are shown over the first 5000 sec for raw fluorescence or 500 sec for normalized values for refolding, although the data were collected over 2 h.

Table 3.

Refolding kinetic rate constants

k1 t1/2 k2 t1/2
Wild type 0.047 15 sec 0.0036 190 sec
W42-only NA NA 0.0036 190 sec
W68-only NA NA 0.0033 210 sec
W130-only 0.022 30 sec 0.0023 300 sec
W156-only 0.022 30 sec 0.0047 150 sec
Open in a new tab

Figure 7A ▶ shows the changes in the raw fluorescence signal during refolding for the four triple-tryptophan mutants. The major transitions occurred within the first 500 sec of refolding. For ease of visual comparison, Figure 7B ▶ shows the refolding fluorescence signals normalized between the denatured and native states. Intermediates formed within the first second of refolding were not detected in these experiments.

Wild-type HγD-Crys was best fit to a three-state model, suggesting the presence of one partially folded intermediate along the productive refolding pathway (Fersht 1999). The refolding reaction exhibited an early transition from denatured to a partially refolded intermediate with a t1/2 of 15 sec, and a second transition of intermediate to native with a t1/2 of 190 sec. The intermediate was more fluorescent than the denatured protein, and the native state was more fluorescent than the intermediate.

graphic file with name M2.gif (2)

The fluorescence signals of the domain I tryptophans reached a native-like state more slowly than the signals from the domain II tryptophans. The fluorescence of unfolding curves observed for W42-only and W68-only could be best fit to one exponential, suggesting a two-state refolding pathway. These chains refolded with t1/2 values of 190 and 210 sec, respectively, and neither populated an observable intermediate. The refolding transitions of these domain I tryptophans correspond closely with the second transition observed in wild type. The domain I constructs do not account for the burst fluorescence kinetics observed in wild type.

The domain II tryptophan constructs, W130-only and W156-only, underwent a refolding process that was initially faster than domain I. These two proteins had fluorescence refolding signals that were best fit to a three-state model, suggesting the existence of an intermediate. Both had initial t1/2 values that were similar to the transitions observed for wild type. W130-only had an initial t1/2 of 27 sec and a secondary t1/2 of 300 sec, whereas the transitions of W156-only had values of 31 and 150 sec, respectively.

W130-only had a hyperfluorescent intermediate, and was the only construct displaying a significant transformation within the dead-time of these experiments as shown by a very high initial fluorescence signal upon the onset of re-folding (Fig. 7A ▶). It is possible that a millisecond refolding intermediate was populated by W130-only that was not readily visible using the syringe-injection port system. Further studies using stopped-flow devices should be performed on this protein to assess the significance of this putative early transformation in the refolding pathway.

From these data, it appeared that domain II refolded first and was then followed by refolding of domain I (Fig. 7B ▶). A global hydrophobic collapse likely occurred in domain II that was then slowly followed by tight packing of the domain II tryptophans into their proper orientation. This resulted in the population of an intermediate with a primarily intact domain II, but a denatured domain I. While Trp130 and Trp156 were being tightly packed, a nearly simultaneous hydrophobic collapse of domain I occurred, as evidenced by the change in fluorescence of Trp 42 and Trp 68. This two-step process resulted in the formation of a protein with native-like tertiary structure as seen by a stabilization in tryptophan fluorescence emission.

Discussion

HγD-Crys is a highly stable protein containing four trypto-phans buried within two highly hydrophobic protein cores. Four stable, triple mutants of His-tagged HγD-Crys were produced, each containing one native tryptophan and three tryptophan-to-phenylalanine substitutions. All of the mutant proteins were expressed in E. coli in the soluble fraction of the cell lysate. In addition, each had a primarily β-sheet character by circular dichroism measurements and displayed bands of a similar mobility as wild type on a native polyacrylamide gel.

Equilibrium unfolding and refolding experiments in Gd-nHCl demonstrated that the triple-mutant tryptophan constructs were destabilized compared with wild type (Fig. 5 ▶). Native to denatured transition midpoints of 1.3 M GdnHCl were measured for W42-only and W68-only N-terminal, domain I and 2.0 M GdnHCl for W130-only and W156-only C-terminal, domain II Wild type had an equilibrium unfolding midpoint of 2.8 M GdnHCl. The decreased transition midpoints of the triple tryptophan mutants are likely due to destabilization of the molecule caused by the triple mutations. Substitution of three tryptophans for phenylalanines probably strains the hydrophobic cores of the molecules, allowing GdnHCl to enter and denature these areas of the protein at lower concentrations than wild type. However, HγD-Crys with single tryptophans substituted by alanine folded into native-like soluble monomers with stabilities equal to that of wild-type HγD-Crys (V. Zepeda, M. Kosin-ski-Collins, S. Flaugh, and J. King, unpubl.).

The triple-tryptophan mutants display no increased or decreased propensity for aggregation with respect to wild-type HγD-Crys during either refolding or long-term storage at 4°C. The overall solubility and native-like structure of HγD-Crys with phenylalanines or alanines substituted for tryptophans suggests that the tryptophan residues are not critical for stability at physiological temperature. Given the location of these substitutions in the hydrophobic core of HγD-Crys, the high solubilities and native-like structures of all mutant tryptophan constructs are surprising.

All four tryptophans in HγD-Crys show high chemical conservation throughout the known members of the γ-crystallin family. Trp 42 and Trp 130 are 100% conserved throughout the γ-crystallin family as aromatic residues, whereas Trp 68 and Trp 156 are 80% and 89% conserved in aromaticity, respectively. The deviations in aromatic composition in Trp 68 and Trp 156 are found in primarily aquatic animals such as carp and catfish. It is interesting to note that the eyes of these animals are shielded from direct light by their underwater surroundings.

The quenching of tryptophan emission in the native state

On the basis of fluorescence spectra data collected from the triple-mutant tryptophan proteins, it appears that Trp 68 and Trp 156 are responsible for the native-state quenching phenomenon in HγD-Crys. Trp 68 is in domain I and Trp 156 is in domain II. Examination of the three-dimensional structure of HγD-Crys revealed that both are located in similar positions within the domains.

Because both Trp 68 and Trp 156 exhibit anomalous quenching, it seemed likely that the residues responsible for this reaction would be similar in the two domains, given the high sequence similarity between domain I and domain II of HγD-Crys. Inspection of the residues within 8 Å of Trp 68 and Trp 156 showed that the only potential quenchers that were constant between the two domains were two tyrosine residues and histidine. Tyr 55, Tyr 62, and His 65 make a “cage” around Trp 68 (Fig. 8A ▶), whereas Tyr 143, Tyr 150, and His 122 surround Trp 156 (Fig. 8B ▶). Chen and Barkley (1998) have suggested that tyrosine may quench tryptophan fluorescence via a proton-transfer mechanism, whereas his-tidine may quench via excited-state electron transfer. As histidine quenches primarily when protonated, and as these experiments were performed at pH 7.0, tyrosine is the likely side chain participating in the quenching phenomenon.

Figure 8.

Figure 8.

Open in a new tab

X-ray structure of Tyr-His-Tyr aromatic “cage” surrounding Trp 68 (A) and Trp 156 (B). Trp 68 is enclosed by Tyr 55, Tyr 62, and His 65, whereas Tyr 143, Tyr 150, and His 122 surround Trp 156 (Basak et al. 2003).

Both Trp 42 and Trp 130 are located within quenching distance of cysteine and histidine side chains. In the original report of the crystal structure of bovine γB crystallin, Wistow et al. (1983) suggested that these interactions may have been the source of the anomalous quenching. Although the results reported here make that mechanism less likely, the cysteine and histidine residues surrounding Trp 42 and Trp 130 may protect these tryptophans from photo damage in vivo.

Unfolding and refolding intermediates

Kinetic analysis of HγD-Crys revealed that unfolding and productive refolding involve a similar intermediate state in which tertiary structure is absent from domain I, but present in domain II (Fig. 9 ▶). The population of this single domain conformer may represent an intermediate important in the in vitro aggregation pathway previously described (Kosinski-Collins and King 2003). Structurally distinct folding intermediates have been shown to be important in many aggregation pathways and disease systems (Haase-Pettingell and King 1988; Wetzel 1994; Speed et al. 1995). The domain-swapping model would provide a mechanism to explain polymerization of such partially folded two-domain species into an ordered fibrillar state (Liu and Eisenberg 2002; Rousseau et al. 2003). Additionally, such partially folded species may be related to the crystallin conformers recognized by α-crystallin (Das et al. 1999; Cobb and Petrash 2002).

Figure 9.

Figure 9.

Open in a new tab

Model of HγD-Crys folding and aggregation. Upon rapid dilution into refolding buffer, denatured HγD-Crys had a putative intermediate that had domain II folded and domain I unfolded. During unfolding in GdnHCl, domain I unfolded earlier than domain II, likely populating a similar intermediate.

These data support a sequential model for tertiary structure formation of the domain cores. However, we do not yet know what contributions interface residues make in the folding of HγD-Crys. Future studies will investigate the kinetic and thermodynamic significance of domain interface residues in folding of the molecule.

Tryptophan quenching and protection from ultraviolet radiation

Lens proteins are subject to irradiation in the visible and ultraviolet light range during the entire human lifetime. The cornea of the eye absorbs the majority of light at wavelengths <295 nm, but allows limited ultraviolet light at wavelengths longer than 295 nm to pass through to the lens and to the crystallin proteins (Sliney 2002). Although the peak of tryptophan absorbance occurs at 278 nm, as with most tryptophan-containing proteins, HγD-Crys shows an absorption tail that extends well beyond 300 nm (Fig. 3A ▶).

It is not obvious why the tryptophan residues are conserved in the γ-crystallin family, but these residues are responsible for the major observed ultraviolet light-absorption events of γ-crystallins in vitro. The well-conserved trypto-phans in the γ-crystallins may have been maintained during evolution as a part of the mechanism of protecting the retina and other eye structures from UV-B damage. However, this would then render the γ-crystallin proteins themselves sensitive to ultraviolet light damage. In epidemiological studies, excess UV-B exposure has been shown to be directly correlated with an increase in mature-onset cataract in humans (Taylor et al. 1988; Sliney 2002). More specifically, photo oxidation of tryptophan residues is known to be a precursor in the formation of brunescent cataracts, and is thought to occur as a result of prolonged exposure to ultraviolet light (Pirie 1971; Kurzel et al. 1973; Zigman et al. 1973; Davies and Truscott 2001; Soderberg et al. 2002). To avoid covalent modification associated with extended exposure of tryptophan residues to ultraviolet light, the crystallins would require a means to dissipate the absorbed excited state energy.

HγD-Crys was more fluorescent in its native state than in its denatured state (Fig. 4 ▶). Although not emphasized in the literature, γS and γB crystallin show this anomalous quenching of their buried tryptophans as well (Rudolf et al. 1990; Wenk et al. 2000). The existence of native-state quenching of tryptophans in a protein exposed to ultraviolet light, but selected for its stability and solubility, may reflect a role in protecting it from the absorption events. Prolonged exposure to ultraviolet light or other oxidative conditions populated in lens over time by HγD-Crys may generate ring opening and other covalent damage well characterized in other systems (Conti et al. 1988; Balasubramanian et al. 1990; Prinsze et al. 1990). The generation of a charged species within the buried core of the crystallins would be expected to destabilize the protein and cause full or partial unfolding. These species would be candidates for precursors to the aggregated state of the crystallins found in mature-onset cataracts.

Materials and methods

Cloning and site-directed mutagenesis

The human γD crystallin coding sequence had been previously cloned as described (Pande et al. 2001). The gene was excised from the pET3a plasmid and ligated into a pQE.1 plasmid (Qiagen) that added an N-terminal 6-His tag to the protein. The integrity of the HγD-Crys gene was confirmed by sequencing at the facilities of Massachusetts General Hospital.

Four triple-mutant proteins containing three tryptophan-to-phe-nylalanine substitutions at positions 42, 68, 130, and 156 were constructed using a PCR-based mutagenesis procedure (Strata-gene). Sequential mutations were made using complementary primer pairs that altered the Trp codon TGG to Phe TTT (Invit-rogen). The mutations were confirmed by sequencing the region of the resulting plasmids (MGH).

Expression and purification of HγD-Crys

Recombinant HγD-Crys was prepared by pQE.1 plasmid transformation into E. coli M15 (pREP4) cells. Protein production was induced by addition of 1 mM IPTG and allowing 4 h of incubation at 37°C. Cultures were pelleted by centrifugation for 15 min, and cells were resuspended in a 10-mM imidazole, 10-mM Tris, 0.5-M NaCl solution. Cells were lysed using six sequential 20-sec bursts of sonication, followed by 40-sec rest cycles. Lysates were spun at 17,000 RPM for 30 min. The resulting supernatant was then applied to a Ni-NTA column, and protein was eluted using an increasing concentration of 250 mM imidazole, 10 mM Tris, 0.5 NaCl at room temperature. Fractions containing protein were dia-lyzed four times against 4 L of 10 mM ammonium acetate (pH 7.0) for 4 h. Maldi-mass spectroscopy was performed on all proteins to confirm the presence of the desired amino acid substitutions (MIT-Biopolymers Lab). All proteins were transferred from ammonium acetate to S Buffer (10 mM NaPO4, 5 mM DTT, 1 mM EDTA at pH 7.0) by dilution.

Circular dichroism

CD spectra of wild-type and mutant HγD-Crys proteins were collected on an Aviv Associates model 202 circular dichroism spectrometer. All readings were performed on 0.3-mg/mL HγD-Crys protein samples in S buffer at 37°C. CD was measured every 1 nm between 200 and 260 nm. The signals at all wavelengths were averaged over 5 sec.

Ultraviolet light absorbance

Ultraviolet light spectra of wild-type and mutant HγD-Crys proteins were collected on a Varian Cary 50 Bio ultraviolet light spectrometer. Readings in the peptide backbone region (190 to 240 nm) were performed on 0.1-mg/mL HγD-Crys in S buffer. Absorbance readings in the aromatic region (240 to 340 nm) were taken of 0.3-mg/mL protein samples in S buffer (native) or S buffer containing 5.5 M GdnHCl (denatured). Protein concentration was calculated by measuring denatured protein absorbance at 280 nm and using a protein-extinction coefficient of 41.04 mM−1 for wild-type and 23.97 mM−1 for triple-mutant tryptophan His-tagged constructs, respectively.

Fluorescence emission spectra

Fluorescence emission spectra were read on a Hitachi F-4500 fluo-rimeter with a continuous flow temperature-control system. Proteins were diluted to a concentration of 10 μg/mL in S buffer or S buffer containing 5.5 M GdnHCl. Samples were excited at 295 nm, and emission was measured from 310 to 420 nm. The excitation and emission slit widths were both set to 10 nm. The background fluorescence of S buffer or S buffer and 5.5 M GdnHCl was subtracted out from the sample reading. Fluorescence emission maxima were calculated by averaging signals over every 5 nm and selecting the midpoint of the five signals that exhibited the highest average.

Equilibrium refolding and unfolding

For the unfolding process, purified wild-type or mutant HγD-Crys was diluted to 10 μg/mL in increasing amounts of GdnHCl in S buffer from 0 to 5.5 M. The samples were incubated at 37°C until equilibrium was reached (about 6 h). For the refolding titration, 100 μg/mL protein was denatured in 5.5 M GdnHCl in S buffer at 37°C for 5 h. The protein was subsequently refolded by dilution to 10 μg/mL into decreasing concentrations of GdnHCl from 5.5 to 0.55 M. The fluorescence spectra of the equilibrated samples were determined using a Hitachi 4500 fluorimeter equipped with a continuous temperature-control system with an excitation wavelength at 295 nm, and emission monitored from 310 to 420 nm. The excitation and emission slits were both set to 10 nm. The ratios of emission intensities of 360 over 320 nm were used for data analysis of wild type, W68-only, and W156-only, W42-only, and W130-only. Fraction unfolded values were calculated using the method of Pace et al. (1989), and denaturation midpoints were calculated using the Kaliedagraph (Synergy Software) curve-fitting function.

Unfolding fluorescence kinetics

Tryptophan environment changes with refolding were monitored using a Hitachi 4500 fluorimeter equipped with a continuous temperature-control system. Native protein (100 μg/mL in S buffer at 37°C) was unfolded by dilution into S buffer to final concentrations of 10 μg/mL HγD-Crys and 5.5 M GdnHCl using a syringe port-injection system exhibiting a dead-time of 1 sec. Loss of global structure was monitored with continuous excitation at 295 nm at 37°C for 1 h. Emission intensities during kinetic unfolding were collected at 350 nm for wild type, W68-only, and W156-only, whereas the emission intensities were collected at 320 nm for W42-only and W130-only. The fluorescence curves were fit to a series of consecutive first-order exponentials using the method described by Fersht (1999). The signals were all fit using the Kaliedagraph (Synergy Software) curve-fitting algorithm to mechanisms having one, two, and three exponentials, and the best fit was selected by inspection. All proteins were unfolded in at least two separate experiments to ensure the accuracy of the observed fluorescence and curve fitting, given the high levels of noise.

Refolding fluorescence kinetics

Changes in tryptophan environment during refolding were monitored using a Hitachi 4500 fluorimeter equipped with a continuous temperature-control system. Native protein was denatured at 100 μg/mL in 5.5 M GdnHCl in S buffer at 37°C for 2 h. The unfolded protein was refolded by dilution into S buffer to final concentrations of 10 μg/mL HγD-Crys and 1.0 M GdnHCl using a syringe port-injection system with a dead-time of 1 sec. Increase in global structure during refolding was monitored with continuous excitation at 295 nm and emission at 350 nm for wild type, W68-only, and W156-only, and emission at 320 nm for W42-only and W130-only. The fluorescence curves were fit to a series of consecutive first-order exponentials using the method described by Fersht (1999). The signals were all fit using the Kaliedagraph (Synergy Software) curve-fitting algorithm to mechanisms having one, two, and three exponentials, and the best fit was selected by inspection. These experiments were repeated for each protein refolding to 1.0 and 1.5 M GdnHCl.

Acknowledgments

We thank Veronica Zepeda and Dr. Peter Weigele for their helpful discussions and technical assistance. This research was supported by NIH grant GM17980, awarded to J.K.

The publication costs of this article were defrayed in part by payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 USC section 1734 solely to indicate this fact.

Article and publication are at http://www.proteinscience.org/cgi/doi/10.1110/ps.04627004.

References

  1. Balasubramanian, D., Du, X., and Zigler, J.S.J. 1990. The reaction of singlet oxygen with proteins, with special reference to crystallins. Photochem. Pho-tobiol. 52 761–768. [DOI] [PubMed] [Google Scholar]
  2. Basak, A., Bateman, O., Slingsby, C., Pande, A., Asherie, N., Ogun, O., Benedek, G.B., and Pande, J. 2003. High resolution x-ray crystal structures of human γD crystallin (1.25 Å) and the R58H mutant (1.15 Å) associated with aculeiform cataract. J. Mol. Biol. 328 1137–1147. [DOI] [PubMed] [Google Scholar]
  3. Bron, A.J., Vrensen, G.F., Koretz, J., Maraini, G., and Harding, J.J. 2000. The ageing lens. Ophthalmologica 214 86–104. [DOI] [PubMed] [Google Scholar]
  4. Chen, Y. and Barkley, M.D. 1998. Towards understanding tryptophan fluorescence in proteins. Biochemistry 37 9976–9982. [DOI] [PubMed] [Google Scholar]
  5. Cobb, B.A. and Petrash, J.M. 2002. α crystallin chaperone-like activity and membrane binding in age-related cataracts. Biochemistry 41 483–490. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Conti, F., Cantu, A.M., and Duclohier, H. 1988. Orientation of the tryptophans responsible for the photoinactivation of nerve sodium channels. Eur. Bio-phys. J. 16 73–81. [DOI] [PubMed] [Google Scholar]
  7. Das, K.P., Choo-Smith, L.-P., Petrash, J.M., and Surewicz, W.K. 1999. Insight into the secondary structure of non-native proteins bound to a molecular chaperone α-crystallin. J. Biol. Chem. 274 33209–33212. [DOI] [PubMed] [Google Scholar]
  8. Davies, M.J. and Truscott, R.J.W. 2001. Photo-oxidation of proteins and its role in cataractogenesis. J. Photochem. Photobiol. 63 114–125. [DOI] [PubMed] [Google Scholar]
  9. Eyles, S.J. and Gierasch, L.M. 2000. Multiple roles of prolyl residues in structure and folding. J. Mol. Biol. 301 737–747. [DOI] [PubMed] [Google Scholar]
  10. Fersht, A. 1999. Structure and mechanism in protein science: A guide to enzyme catalysis and protein folding. W.H. Freeman and Company, New York.
  11. Fu, L. and Liang, J. 2003. Alteration of protein-protein interactions of congenital cataract crystallin mutants. Invest. Ophthalmol. Vis. Sci. 44 1155–1159. [DOI] [PubMed] [Google Scholar]
  12. Haase-Pettingell, C.A. and King J. 1988. Formation of aggregates from a thermolabile folding intermediate in P22 tailspike maturation. A model for inclusion body formation. J. Biol. Chem. 263 4977–4983. [PubMed] [Google Scholar]
  13. Hanson, S.R., Hasan, A., Smith, D.L., and Smith, J.B. 2000. The major in vivo modifications of the human water-insoluble lens crystallins are disulfide bonds, deamidation, methionine oxidation and backbone cleavage. Exp. Eye Res. 71 195–207. [DOI] [PubMed] [Google Scholar]
  14. He, Q., Mason, A.B., Lyons, B.A., Tam, B.M., Nguyen, V., MacGillivray, R.T.A., and Woodworth, R.C. 2001. Spectral and metal-binding properties of three single-point tryptophan mutants of the human transferrin N-lobe. Biochem. J. 354 423–429. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Heon, E., Priston, M., Schorderet, D.F., Billingsley, G.D., Girard, P.O., Lubsen, N., and Munier, F.L. 1999. The γ-crystallins and human cataracts: A puzzle made clearer. Am. J. Hum. Genet. 65 1261–1267. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Jaenicke, R. 1999. Stability and folding of domain proteins. Prog. Biophys. Mol. Biol. 71 155–241. [DOI] [PubMed] [Google Scholar]
  17. Kosinski-Collins, M.S. and King, J. 2003. In vitro unfolding, refolding, and polymerization of human γD crystallin, a protein involved in cataract formation. Protein Sci. 12 480–490. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kurzel, R.B., Wolbarsht, M., Yamanashi, B.S., Staton, G.W., and Borkman, R.F. 1973. Tryptophan excited states and cataracts in the human lens. Nature 241 132–133. [DOI] [PubMed] [Google Scholar]
  19. Lakowicz, J.R. 1999. rinciples of fluorescence spectroscopy. Kluwer Academic Plenum Publishers, New York.
  20. Liu, Y. and Eisenberg, D. 2002. γD domain swapping: As domains continue to swap. Protein Sci. 11 1285–1299. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Mayr, E.M., Jaenicke, R., and Glockshuber, R. 1997. The domains in γB crystallin: Identical fold—different stabilities. J. Mol. Biol. 269 260–269. [DOI] [PubMed] [Google Scholar]
  22. McCarty, C.A. and Taylor, H.R. 2002. A review of the epidemiological evidence linking ultraviolet radiation and cataracts. In rogress in lens and cataract research (eds. O. Hockwin et al.). Dev. Ophthalmol. 35 21–31. [DOI] [PubMed] [Google Scholar]
  23. Nandrot, E., Slingsby, C., Basak, A., Cherif-Chefchaouni, M., Benazzouz, B., Hajaji, Y., Boutayeb, S., Gribouval, O., Arbogast, L., Berraho, A., et al. 2003. γ-D crystallin gene (CRYGD) mutation causes autosomal dominant congenital cerulean cataracts. J. Med. Genet. 40 262–267. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. National Eye Institute (US) 2002. “Vision problems in the US.” Prevent Blindness in America.
  25. Oyster, C.W. 1999. The human eye structure and function. Sinauer Associates, Inc., Sunderland, MA.
  26. Pace, C.N., Shirley, B.A., and Thomson, J.A. 1989. Measuring the conformational stability of a protein. In rotein structure: A practical approach (ed. T.E. Creighton), pp. 311–330. Oxford University Press, UK.
  27. Pande, A., Pande, J., Asherie, N., Lomakin, A., Ogun, O., King, J.A., and Benedek, G.B. 2001. Crystal cataracts: Human genetic cataract caused by protein crystallization. Proc. Natl. Acad. Sci. 98 6116–6120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Pirie, A. 1971. Formation of N-formylkynurenine in proteins from lens and other sources by exposure to sunlight. Biochem J. 125 203–208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Prinsze, C., Dubbleman, T.M., and Van Steveninck, J. 1990. Protein damage induced by small amounts of photodynamically generated singlet oxygen or hydroxyl radicals. Biochim. Biophys. Acta. 1038 152–157. [DOI] [PubMed] [Google Scholar]
  30. Rousseau, F., Schymkowitz, J.W., and Itzhaki, L.S. 2003. The unfolding story of three-dimensional domain swapping. Structure 11 243–251. [DOI] [PubMed] [Google Scholar]
  31. Rudolph, R., Sienbendritt, R., Nesslauer, G., Sharma, A.K., and Jaenicke, R. 1990. Folding of an all-β protein: Independent domain folding in γII-crystallin from calf eye lens. Proc. Natl. Acad. Sci. 87 4625–4629. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Sasaki, H., Jonasson, F., Shui, Y.B., Kojima, M., Ono, M., Katoh, N., Cheng, H.-M., Takahashi, N., and Sasaki, K. 2002. High prevalence of nuclear cataract in the population of tropical and subtropical areas. In rogress in lens and cataract research (eds. O. Hockwin et al.). Dev. Ophthalmol. 35 60–69. [DOI] [PubMed] [Google Scholar]
  33. Sliney, D.H. 2002. How light reaches the eye and its components. Inter. J. Toxicol. 21 501–509. [DOI] [PubMed] [Google Scholar]
  34. Soderberg, P.G., Lofgren, S., Ayala, M., Dong, X., Kakar, M., and Mody, V. 2002. Toxicity of ultraviolet radiation exposure to the lens expressed by maximum tolerable dose. In rogress in lens and cataract research (eds. O. Hockwin et al.). Dev. Ophthalmol. 35 70–75. [DOI] [PubMed] [Google Scholar]
  35. Speed, M.A., Wang, D.I.C., and King, J. 1995. Multimeric intermediates in the pathway to the aggregated inclusion body state for P22 tailspike polypeptide chains. Protein Sci. 4 900–908. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Taylor, H.R., West, S.K., Rosenthal, F.S., Munoz, B., Newland, H.S., Abbey, H., and Emmet, E.A. 1988. Effects of ultraviolet radiation on cataract formation. N. Engl. J. Med. 319 1429–1433. [DOI] [PubMed] [Google Scholar]
  37. Wenk, M., Herbst, R., Hoeger, D., Kretschmar, M., Lubsen, N.H., and Jaenicke, R. 2000. γ S-crystallin of bovine and human eye lens: Solution structure, stability and folding of the intact two-domain protein and its separate domains. Biophys. Chem. 86 95–108. [DOI] [PubMed] [Google Scholar]
  38. Wetzel, R. 1994. Mutations and off-pathway aggregation of proteins. Trends Biotechnol. 12 193–198. [DOI] [PubMed] [Google Scholar]
  39. Wistow, G., Turnell, B., Summers, L., Slingsby, C., Moss, D., Miller, L., Lind-ley, P., and Blundell, T. 1983. X-ray analysis of the eye lens protein γ-II crystallin at 1.9 Å resolution. J. Mol. Biol. 170 175–202. [DOI] [PubMed] [Google Scholar]
  40. Zigman, S., Schultz, J., and Yulo, T. 1973. Possible roles of near ultraviolet light in the cataractous process. Exp. Eye Res. 15 201. [DOI] [PubMed] [Google Scholar]

Articles from Protein Science : A Publication of the Protein Society are provided here courtesy of The Protein Society

RESOURCES