Biomacromolecules 2004, 5, 1362-1370
1362
Guanidine Hydrochloride Can Induce Amyloid Fibril Formation from Hen Egg-White Lysozyme Brian A. Vernaglia,* Jia Huang, and Eliana D. Clark† Department of Chemical and Biological Engineering, Tufts University, Medford, Massachusetts 02155 Received February 17, 2004; Revised Manuscript Received April 29, 2004
The formation of amyloid fibrils is an intractable problem in which normally soluble protein polymerizes and forms insoluble ordered aggregates. Such aggregates can range from being a nuisance in vitro to being toxic in vivo. The latter is true for lysozyme, which has been shown to form toxic deposits in humans. In the present study, the effects of partial denaturation of hen egg-white lysozyme via incubation in a concentrated solution of the denaturant guanidine hydrochloride are investigated. Results show that when lysozyme is incubated under moderate guanidine hydrochloride concentrations (i.e., 2-5 M), where lysozyme is partially unfolded, fibrils form rapidly. Thioflavin T, Congo red, X-ray diffraction, transmission electron microscopy, atomic force microscopy, and circular dichroism spectroscopy are all used to verify the production of fibrils under these conditions. Incubation at very low or very high guanidine hydrochloride concentrations fails to produce fibrils. At very low denaturant concentrations, the structure of lysozyme is fully native and very stable. On the other hand, at very high denaturant concentrations, guanidine hydrochloride is capable of dissolving and dis-aggregating fibrils that are formed. Raising the temperature and/or concentration of lysozyme accelerates fibril formation by further adding to the concentration of partially unfolded species. The addition of preformed fibrils also accelerates fibril formation but only under partially unfolding conditions. The results presented here provide further evidence that partial unfolding is a prerequisite to fibril formation. Partial denaturation can accelerate fibril formation in much the same way that mutations have been shown to accelerate fibril formation. Introduction Amyloid fibrils are highly ordered structures in which individual protein molecules aggregate and are aligned with each other in a long-range repetitive manner, much like a crystal. Numerous proteins have been identified as forming amyloid,1 and at least 18 different clinical syndromes related to amyloid have been reported in the literature.2 Although each protein has a unique and characteristic native fold and distinct amino acid sequences, the fibrils that they form share morphological features.3 Mature fibrils are between 60 and 120 Å in width and have indefinite lengths. They are typically composed of between 2 and 5 smaller protofilaments. The protofilaments themselves are composed of cross β-sheet structures with β-strands perpendicular to the fibril axis and backbone hydrogen bonds parallel to the axis and therefore commonly possess the β-sheet spacing of approximately 4.7 Å. This spacing is visible in X-ray diffraction images of fibrils.1,4 Due to the morphological similarities between many fibrils from different protein building blocks, it has been hypothesized that different proteins follow similar fibril formation pathways.3 Most studies suggest that proteins must unfold from a native tertiary (monomeric) structure to a partially * To whom correspondence should be addressed. Address: 4 Colby Street, Medford, MA 02155. Telephone: 617-627-3900. Fax: 617-6273991. Email:
[email protected]. † Present address: Genzyme Corp., 76 New York Ave., Framingham, MA 01701.
unfolded (monomeric) intermediate. In the case of native multimeric proteins, most studies suggest that dissociation of the multimer to a monomer prior to partial unfolding is also a prerequisite. The partially unfolded intermediate may undergo further structural rearrangement and start forming fibrils. Processes that tend to favor the partially denatured intermediate species tend to speed up this transformation. Heat,5 low pH,6 agitation,7,8 and pressure9 have all been shown to induce or speed up fibrillation. Molecular crowding (with polymers that simulate cytoplasmic proteins) has also been shown to speed up the fibrillation process10 Fibril formation is a nucleation-polymerization event: no aggregation occurs for protein concentrations below a critical value and once the protein concentration exceeds that critical value by a small amount, there is a lag time before polymerization occurs.11 Two parameters, lag time and growth rate, are sufficient to describe the sigmoidal growth that follows.12 Hen egg-white lysozyme (lysozyme) is a monomeric protein composed of 129 amino acids and has a molecular weight of 14.3 kDa. The structure of hen egg-white lysozyme was the first determined by X-ray diffraction13 and is primarily R-helical (∼30%) with two short β-strands (∼6%).14 There are eight cysteines present in the primary structure of lysozyme, resulting in four disulfide bonds. There is a high degree of sequence and structural homology between hen egg white and human lysozyme. In 1993, lysozyme was found to form fibrils.15 Specifically, it was shown that the
10.1021/bm0498979 CCC: $27.50 © 2004 American Chemical Society Published on Web 06/19/2004
Guanidine Hydrochloride Induced Fibril Formation
single point mutations of human lysozyme I56T and D67H were present in fibril deposits in affected patients. Subsequently, it was shown that wild type (wt) human lysozyme could form fibrils in vitro. Work from Dobson’s group has shown that lowering the pH to 2.0 and raising the temperature to 57 °C result in amyloid fibrils.16 They are formed, however, more slowly than the mutant variants (I56T and D67H) under identical conditions. All fibrils formed over a time period of days. They also found that seeding of wt lysozyme with fibrils composed of mutant lysozyme rapidly sped up the fibrillation process. However, ex vivo amyloid samples from patients do not contain wt lysozyme,15 suggesting that the formation process favors the less stable mutant species in vivo. Dobson’s group found similar results for wt hen egg-white lysozyme, in which heating at low pH resulted in the formation of amyloid fibrils.17 Again, they found that seeding wt lysozyme with fully formed fibrils accelerated the fibril formation process. The work of Goda and co-workers18 indicated that wt hen egg-white lysozyme fibrils could also be induced by subjecting lysozyme to highly concentrated ethanol solutions. At 90% ethanol concentration, the percentage of β-sheet content rises from 6 to 26%. Upon incubation, fibrils formed as identified by electron microscopy and Congo red staining. While studying insulin fibril formation, rather than that of lysozyme, Nielsen and co-workers12 showed that, as the concentration of the denaturant urea was increased from 0 to 2 to 4 to 6 M, lag times before growth decreased and growth rates increased. They concluded that urea promoted a population of partially folded species that decreased lag times and increased growth rates. GdnHCl is a strong denaturant which can coat the exterior of proteins. Above certain protein-specific concentrations, GdnHCl can fully denature a protein.19 While denatured, however, disulfide-containing proteins are never fully unfolded even by high levels of GdnHCl. A considerable degree of ordered structure remains due to the presence of disulfide bonds, which GdnHCl does not affect.20 Low levels of GdnHCl (1 M), lysozyme is at least partially denatured. The inverse is also true: fibrils never form under conditions where lysozyme possesses 100% native structure.
It appears that a disruption of the native protein structure is a prerequisite for lysozyme fibrils formation. Upon addition of moderate levels of GdnHCl, which disrupt the structure, fibrils form rapidly. The question remains as to why higher levels of GdnHCl slow the fibril formation process. One explanation would be that fully denatured lysozyme cannot form fibrils. However, as revealed in Figures 1 and 8, fibrils form in 4 and 5 M GdnHCl at 50 °C where lysozyme is almost fully or fully denatured. CD results in Figure 10 verify that lysozyme is fully native at 1 M GdnHCl but starts to lose structure above 1 M GdnHCl. At 4 and 5 M GdnHCl, lysozyme is fully denatured. Even fully denatured lysozyme, however, has substantial residual structure due to the presence of four disulfide bonds. Thus, full denaturation of lysozyme still only results in partial unfolding, and it does not appear to be adequate for preventing fibril formation. The answer likely lies in GdnHCl’s ability to dissolve fibrils at higher concentrations. To test this hypothesis, aliquots of mature fibril suspensions were centrifuged at 14 100 g for 20 min at room temperature and resuspended in various concentrations of GdnHCl ranging from 0 to 8 M. After incubation for 1 h, these samples were again centrifuged at 14 100 g for 20 min at room temperature, and the fraction of protein present in the supernatant, which is a measure of how much protein is resolubilized from fully formed fibrils by GdnHCl, was determined. Figure 11 depicts the quantity of fibrils that can be resolubilized. As can be seen in Figure 11, fibrils in 0-4 M GdnHCl are only minimally dissolved. Above 4 M, fibrils are dissolved to an increasingly greater extent up to 8 M GdnHCl where most of the lysozyme fibrils are dissolved. Of course, this experiment differs dramatically from the fibril forming experiments taking place where nascent fibril nuclei and proto-fibrils are exposed to the denaturant. Under those conditions, GdnHCl may be even more effective in dissolving fibrils and fibril precursors or preventing them from aggregating. Effects of Temperature, Concentration, and Seeding on Lysozyme Fibril Formation. Although the focus of this work is on the effects of the denaturant GdnHCl, in the absence of adequate temperature, concentration, and mixing, the formation of fibrils is painstakingly slow. An examination of these factors can help further define the role of GdnHCl in fibril formation.
1368
Biomacromolecules, Vol. 5, No. 4, 2004
Figure 11. Lysozyme resolubilized by GdnHCl. Low levels are resolubilized from 0 to 4 M GdnHCl. Above 4 M GdnHCl, an increasingly greater percent of total fibrils is resolubilized. Fibrils were produced in 3 M GdnHCl (pH 6.3, 50 °C). Samples were centrifuged to collect fibrils, resuspended in various concentrations of GdnHCl, and then recentrifuged. The resulting supernatant was analyzed.
Vernaglia et al.
affected. The baseline is successively higher with increasing seeding levels, indicating the ThT binding to the seeds themselves. Fibrils did not form under any seeding conditions in 0 or 6 M GdnHCl (data not shown). At 0 M, the elevation of the baseline was present again due to the presence of the seeds. At 6 M, the baselines were identical under all seeding conditions, indicating that GdnHCl dissolved the fibrils during the delay between loading the plate and analysis. The successful formation of fibrils through seeding at 3 M GdnHCl, combined with the failure to produce fibrils through seeding at both 0 and 6 M GdnHCl again suggests that, although the presence of seed fibrils accelerates the formation of lysozyme fibrils, the presence of moderate levels of GdnHCl is also necessary to partially unfold lysozyme and make it amenable to forming fibrils. Discussion
Figure 12. Effects of seeding on the formation of lysozyme fibrils in a microtiter plate. Fibril samples (0, 0.005, 0.020, and 0.080 mg/mL) were added to samples of native lysozyme (2 mg/mL) resulting in various weight percents of fibrillar to native lysozyme (0, 0.25, 1, and 4 wt %). 20 mM ThT was added to each sample. The GdnHCl concentration was adjusted to 3 M, and samples were incubated at 45 °C in a fluorescence microtiter plate reader. Samples were agitated every 10 min for 20 s.
Temperature plays a dual role in fibril formation in that higher temperature can both partly unfold lysozyme and can increase reaction rates. Under identical conditions (1 or 5 mg/mL, 2.66 M GdnHCl), increasing the temperature from 40 to 50 °C can dramatically increase γ and reduce t. Likewise, increasing the concentration of protein speeds up fibril formation, although to a lesser extent than temperature. Specifically, an increase of T from 40 to 50 °C results in a 15-fold increase in γ and a 10-fold decrease in t. An increase of lysozyme concentration from 1 to 5 mg/mL results in a 3-fold increase in γ and only a slight decrease in t. Finally, samples of lysozyme were incubated in a microtiterplate at 45 °C with 0, 3, and 6 M GdnHCl in the presence of 0.25, 1, and 4 weight % seeds of already formed fibrils. The use of a microtiterplate (rather than a stirred cuvette) slows down the overall fibril formation process such that fibrils are not seen under any GdnHCl concentration over the course of ∼80 h in the absence of seeding. The results, depicted in Figure 12, reveal that the lag time is dramatically reduced by the addition of preformed fibrils only in the samples in 3 M GdnHCl when the seeding exceeded 0.25%. Growth rates, as indicated by slopes, were not significantly
With increasing concentrations of GdnHCl, lysozyme is partially denatured, and in the presence of heating and stirring, fibrillation takes place. In zone I [Figure 2, 0-1 M GdnHCl], mild agitation with a Teflon stirring bar, heating to 50 °C, and low levels of GdnHCl do not promote fibril production. Under these conditions, lysozyme is natively folded. Low levels of amorphous aggregation can take place under these conditions. These amorphous aggregates do not appear to act as seeds for forming fibrils. Furthermore, the addition of mature fibrils does not induce fibrillation. Upon an increase in GdnHCl concentration [zone II; 2-3 M GdnHCl], a dramatic increase in fibrillation takes place. The effect of GdnHCl in this region is to partially unfold lysozyme, changing the conformation to one that can form fibrils more easily. The formation of fibrils forms a positive feedback loop as the fibrils promote the conversion of protein into a pre-fibrillar state. Although GdnHCl likely has some low ability to dissolve fibrils in this zone, the dissolution rate of fibrils is less significant in zone II than the formation rate. The addition of mature fibrils, in concert with the ready supply of partially unfolded lysozyme monomers, rapidly increases the rate of fibrillation (Figure 12). Upon further increase in GdnHCl concentration [zone III; 3-5 M GdnHCl], lysozyme begins to become fully denatured. However, there is significant residual structure due to the presence of disulfide bonds, and even fully denatured lysozyme can still form fibrils as it is still only partially unfolded. Simultaneously, however, GdnHCl is able to easily dissolve fibrils to a moderate extent because of the greater concentration of GdnHCl. Fibrillation decreases in zone III, as GdnHCl readily dissolves fibrils as they are formed. Finally, upon even further increases in GdnHCl concentration [zone IV; 6 M GdnHCl], fibrillation ceases to take place. GdnHCl is too effective at both denaturing lysozyme, perhaps preventing it from forming a stable intermediate conformation with which to form fibrils, and rapidly dissolving newly formed fibrils at such high concentrations. A simple model, such as the one depicted in Figure 13, can suggest the mechanisms taking place. At 0 M GdnHCl,
Guanidine Hydrochloride Induced Fibril Formation
Biomacromolecules, Vol. 5, No. 4, 2004 1369
form remarkable faster than wt species. Our results further support the notion that partial denaturation, either through mutation or environmental changes, are necessary to form fibrils. Summary
Figure 13. Proposed mechanism for lysozyme fibril formation. Open circles are native lysozyme. Filled circles are partially denatured lysozyme monomers. S-curves are fully denatured lysozyme monomers with intact disulfide bonds.
lysozyme is extraordinarily stable, and remains as a native monomer even upon heating to 50 °C, stirring, and the addition of seed fibrils. Upon increasing GdnHCl concentration and moving into zone II (e.g., 3 M GdnHCl), the partially unfolded/partially denatured monomer becomes favored, making fibril production possible. Further increasing GdnHCl concentration to 6 M stops fibril production via two mechanisms: First, the protein is completely denatured (though still only partially unfolded), and perhaps less amenable to forming fibrils. Second, a new pathway whereby fibrils are dissolved by GdnHCl becomes important. Heating, stirring, and seeding all fail to induce fibril formation. The effects of concentration, temperature, and stirring may also be understood in the context of the model in Figure 13, as all three increase the concentration of the partially unfolded monomer either directly or indirectly. Higher initial concentrations of native monomer will in turn result in higher concentrations of partially unfolded monomers as a matter of achieving equilibrium. Increasing the temperature will also result in a higher concentration of partially unfolded monomers, as the equilibrium will also be shifted toward the less stable species. Higher temperature will also result in faster overall kinetics. The addition of fully formed fibrils acts by providing seed fibrils upon which other fibrils can grow but only in the presence of partially unfolded lysozyme monomers. The works of Dobson and co-workers16,17 have shown that heating of wt human and hen egg-white lysozyme could form fibrils, on the time period of several days. Mutations that destabilized the native fold (e.g., human lysozyme I56T) greatly accelerated fibril formation. Still, fibrils from mutant species generally formed on the order of many hours to days. In the present work, we show that partially denatured wt lysozyme can form fibrils at an accelerated rate and form fibrils in a matter of a few hours. Our results compliment those of Dobson and co-workers16,17 and further implicate partial denaturation as a prerequisite to fibril formation both in vitro and in vivo. We are able to mimic the effects of structural mutation by the addition of the chaotropic agent GdnHCl. Other more medically significant proteins, such as transthyretin, form fibrils from both wt and mutant species.6 In the case of transthyretin, mutant species which destabilize the native fold
In the present work, the pathway of natively folded lysozyme to fibrillar aggregate is presented. To examine the role of partial unfolding in fibril formation, the proteins were dissolved in concentrated GdnHCl solutions. The novel results presented detail how GdnHCl is able to affect the tertiary and secondary structure of the native protein and assist in the formation of fibrils. These results provide further evidence that a partially unfolded monomer is the fibril building block for lysozyme. Acknowledgment. The authors thank Regina Valluzzi for assistance with the X-ray diffraction experiments and Lars Waldmann for helpful discussions. References and Notes (1) Dobson, C. M. Trends Biochem. Sci. 1999, 24, 329-332. (2) Sipe, J. D.; Cohen, A. S. J. Struct. Biol. 2000, 130 (2/3), 88-98. (3) Serpell, L. C.; Sunde, M.; Benson, M. D.; Tennent, G. A.; Pepys, M. B.; Fraser, P. E. J. Mol. Biol. 2000, 300 (5), 1033-1039. (4) Rochet, J.-C.; Lansbury, P. Curr. Opin. Struct. Biol. 2000, 10, 6068. (5) Waugh, D. F. J. Am. Chem. Soc. 1946, 68 (2), 247-250. (6) Lai, Z.; Colon, W.; Kelly, J. W. Biochemistry 1996, 35 (20), 647082. (7) Sluzky, V.; Tamada, J. A.; Klibanov, A. M.; Langer, R. Proc. Natl. Acad. Sci. U.S.A. 1991, 88 (21), 9377-81. (8) Sluzky, V.; Klibanov, A. M.; Langer, R. Biotechnol. Bioeng. 1992, 40 (8), 895-903. (9) Ferrao-Gonzales, A. D.; Souto, S. O.; Silva, J. L.; Foguel, D. Proc. Natl. Acad. Sci. U.S.A. 2000, 97 (12), 6445-6450. (10) Shtilerman, M. D.; Ding, T. T.; Lansbury, P. T. Biochemistry 2002, 41, 3855-3860. (11) Harper, J. D.; Lansbury, P. T., Jr. Annu. ReV. Biochem. 1997, 66, 385-407. (12) Nielsen, L.; Khurana, R.; Coats, A.; Frokjaer, S.; Brange, J.; Vyas, S.; Uversky, V. N.; Fink, A. L. Biochemistry 2001, 40 (20), 60366046. (13) Blake, C.; Koenig, D.; Mair, G.; North, A.; Phillips, D.; Sarma, V. Nature 1965, 206, 757-761. (14) Vaney, M. C.; Maignan, S.; RiesKautt, M.; Ducruix, A. Acta Crystallogr.; Sect. D: Biol. Crystallogr. 1996, 52, 505. (15) Pepys, M. B.; Hawkins, P. N.; Booth, D. R.; Vigushin, D. M.; Tennent, G. A.; Soutar, A. K.; Totty, N.; Nguyen, O.; Blake, C. C.; Terry, C. J. Nature 1993, 362 (6420), 553-7. (16) Morozova-Roche, L. A.; Zurdo, J.; Spencer, A.; Noppe, W.; Receveur, V.; Archer, D. B.; Joniau, M.; Dobson, C. M. J. Struct. Biol. 2000, 130 (2/3), 339-351. (17) Krebs, M. R. H.; Wilkins, D. K.; Chung, E. W.; Pitkeathly, M. C.; Chamberlain, A. K.; Zurdo, J.; Robinson, C. V.; Dobson, C. M. J. Mol. Biol. 2000, 300 (3), 541-549. (18) Goda, S.; Takano, K.; Yamagata, Y.; Nagata, R.; Akutsu, H.; Maki, S.; Namba, K.; Yutani, K. Protein Sci. 2000, 9 (2), 369-375. (19) Timasheff, S. Biochemistry 1992, 41, 9857-9864. (20) Hu, C.-H.; Zou, C.-L. Sci. Chin., Ser. B: Chem. 1992, 35 (10), 12141221. (21) Hevehan, D. L.; De Bernardez Clark, E. Biotechnol. Bioeng. 1997, 54, 221-230. (22) Ahmad, A.; Millett, I. S.; Doniach, S.; Uversky, V. N.; Fink, A. L. Biochemistry 2003, 42, 11404-11416. (23) Saxena, V.; Wetlaufer, D. Biochemistry 1970, 9, 5015-5023. (24) Nozaki, Y. Methods Enzymol. 1972, 26, 43-40. (25) LeVine, H. Methods Enzymol 1999, 309, 274-285.
1370
Biomacromolecules, Vol. 5, No. 4, 2004
(26) Klunk, W. E.; Pettegrew, J. W.; Abraham, D. J. J. Histochem. Cytochem. 1989, 37 (8), 1293-7. (27) Klunk, W. E.; Pettegrew, J. W.; Abraham, D. J. J. Histochem. Cytochem. 1989, 37 (8), 1273-81. (28) Klunk, W. E.; Jacob, R. F.; Mason, R. P. Methods Enzymol. 1999, 309, 285-305. (29) Souillac, P. O.; Uversky, V. N.; Millett, I. S.; Khurana, R.; Doniach, S.; Fink, A. L. J. Bio. Chem. 2002, 277 (15), 12666-12679.
Vernaglia et al. (30) Pace, N.; Shirley, B.; Thompson, J. In Protein Structures: A practical approach, 2nd ed.; Creighton, T., Ed.; Oxford University Press: Oxford, U.K., 1996. (31) Bouchard, M.; Zurdo, J.; Nettleton, E. J.; Dobson, C. M.; Robinson, C. V. Protein Sci. 2000, 9 (10), 1960-1967.
BM0498979
