DOI: 10.1002/asia.201500482
Full Paper
Structure Determination
Aggregation of Beta-Amyloid Peptides Proximal to Zwitterionic Lipid Bilayers Chien-I Yang,[a] Brook N. F. Tsai,[a] Shing-Jong Huang,[b] Ting-Yu Wang,[b] Hwan-Ching Tai,*[a] and Jerry C. C. Chan*[a] Abstract: One of the hallmarks of Alzheimer’s disease is the deposition of amyloid plaques, which consist of b-amyloid (Ab) peptides in fibrillar states. Nonfibrillar Ab aggregates have been considered as an important intermediate in the pathway of fibrillization, but little is known about the formation mechanism. The on-pathway b-sheet intermediates of Ab40 peptides can be trapped by incubating the peptides in liposomes formed by zwitterionic lipids. The aggregates of Ab40 peptides have been prepared at a peptide concentra-
Introduction Aggregation of b-amyloid (Ab) peptides has been closely associated with the progression of Alzheimer’s disease (AD).[1–3] On the basis of in vitro studies, the molecular structures of Ab40 fibrils have been largely clarified.[4, 5] In the crowded environment of both intra- and extracellular space, both are full of membrane-enclosed organelles and vesicles. It is unlikely for Ab to aggregate in vivo following the type of kinetics it exhibits in free solution. Interactions between Ab peptides and lipid membranes have been studied extensively to unravel the molecular mechanism underlying the pathogenesis of AD.[6–10] Many studies show that membranes containing anionic lipids strongly affect the molecular structure of Ab fibrils.[7, 10] In particular, it has been demonstrated that liposomes formed by dimyristoylphosphatidylcholine (DMPC)/dimyristoylphosphatidylglycerol (DMPG) can significantly influence the fibrillization kinetics and molecular structure of Ab40 fibrils.[14] On the other hand, freshly solubilized Ab does not interact strongly with phosphatidylcholine (PC) at neutral pH.[7, 11, 12] Furthermore, the major phospholipids in neuronal soma are PC and phosphatidylethanolamine (PE).[13] Therefore, it is of interest to study the [a] C.-I. Yang, B. N. F. Tsai, Prof. Dr. H.-C. Tai, Prof. Dr. J. C. C. Chan Department of Chemistry, National Taiwan University No. 1, Section 4, Roosevelt Road, Taipei 106 (Taiwan) E-mail: [email protected] [email protected] [b] Dr. S.-J. Huang, Dr. T.-Y. Wang Instrumentation Center, National Taiwan University No. 1, Section 4, Roosevelt Road, Taipei 106 (Taiwan) Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/asia.201500482. Chem. Asian J. 2015, 10, 1967 – 1971
tion of less than 10 mm. Solid-state NMR spectroscopy data show that the backbone conformation of the aggregates is almost identical to that of the fibrils formed in free solution. In contrast to anionic lipids, zwitterionic lipids, which are typical of neuronal soma, did not induce any significant conformational difference in Ab40 fibrils. This liposome–Ab system may serve as a useful model to study the fibril formation mechanism.
molecular structure of Ab peptides incubated with neutral liposomes. Herein, we show that the 40-residue b-amyloid peptides (Ab40) incubated in the presence of PC/PE liposomes fold into an intermediate state that has a similar structure to that observed for fibrils incubated in free solution.
Results and Discussion Determination of the Peptide Concentration of LipoAb The approach of reverse-phase evaporation[15] was adopted for the preparation of our liposome–peptide sample, namely, lipoAb, from a 150 mm solution of Ab40 monomer. The peptides neither encapsulated by, nor interacting strongly with, liposomes were removed by size exclusion chromatography (SEC), in which two fractions containing liposomes and peptide monomers were well separated (Figure 1). The integrity of the liposomes was monitored by dynamic light scattering (DLS) measurements. The average diameter was found to be about 250 nm with a polydispersity index (PdI) equal to 0.19 0.02 (Figure S1 in the Supporting Information). The amount of Ab40 peptides in lipoAb was quantified with the dot-blot assay by using the 6E10 antibody.[16] A stock solution of Ab40 peptides (150 mm, incubated for 2 days) was prepared and then diluted with 20 mm Tris buffer to 2, 4, 6, 10, and 20 mm. These standard solutions were used to quantify the amount of Ab40 peptides in lipoAb. From the integrals of the fluorescence signals of the standard solutions (Figure 2), the Ab40 concentration of lipoAb was determined to be (9.1 0.3) mm and that of the monomer fraction was (23.2 0.6) mm, whereas the local concentration of Ab40 within a liposome should be close to 150 mm. On the basis of the amount of lipid and the average
1967
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper
Figure 3. TEM image of the lipoAb-1d sample stained with 2 % uranyl acetate. Intact liposomes were readily observed.
Figure 1. Typical SEC profile for sample purification. The purified lipoAb was collected in the fractions of 37.5–57.0 mL. The free Ab40 monomers were eluted at 72.0–82.5 mL.
Figure 2. Signal integrals of the fluorescence intensities of the dot-blot assay of the lipoAb sample, the two-fold diluted monomer fraction, and standard solutions of Ab40.
mained unruptured was consistent with the notion that zwitterionic lipids had relatively weak interactions with Ab peptides. To characterize the aggregation state of the Ab40 peptides by the thioflavin-T (ThT) assay, the lipoAb samples were treated with 1 % (v/v) Triton X-100 to lyse the liposomes. The mature fibrils of each sample, if any, were collected by centrifugation for subsequent ThT measurements. Figure 4 shows the ThT data of lipoAb samples incubated for different periods. The lipoTris sample, which served as a control, was prepared in the absence of Ab40 peptides. As expected, the ThT intensity of lipoAb-1d would be only marginally higher than that of lipoTris. The ThT intensity increased gradually as the incubation period increased up to 15 days. As another control, the lysed lipoAb-8d sample was incubated for an additional 7 days (lipoAb-8d-lysis-7d). The significant enhancement of the ThT intensity provided a strong indication of fibril formation. The corresponding TEM image also confirmed the formation of mature fibrils (Figure S4 in the Supporting Information). Consequently, we concluded that the majority of the peptides in lipoAb were not in a fibrillar state. Indeed, fibrils were sparsely found in the TEM images of the lipoAb samples. The ThT data
diameter of each liposome, we estimated that at least 26 % of the Ab40 peptides were residing outside the liposomes for lipoAb (see the Supporting Information). Characterization of LipoAb The lipoAb samples were incubated at 37 8C under quiescent conditions for various periods from 2 h to 15 days. The incubated samples are henceforth referred to as lipoAb-2 h, and so on. Figure 3 shows the negative-stain TEM image of the lipoAb-1d sample, in which liposomes were readily found, although a minute amount of fibrils were also observed. As expected, the liposomes were multivesicular, as observed in the cryo-TEM image (Figure S2 in the Supporting Information). DLS measurements confirmed that the liposomes were intact during the whole course of the experiment (Figure S3 in the Supporting Information). The fact that most liposomes reChem. Asian J. 2015, 10, 1967 – 1971
www.chemasianj.org
Figure 4. ThT data of lipoAb samples after lysing by 1 % (v/v) Triton.
1968
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper alone, however, could not elucidate whether the peptides were in monomeric form or in certain intermediate states along the fibrillization pathway. Solid-State NMR Spectroscopy Measurements of LipoAb To probe the molecular structure of the peptides in lipoAb, we prepared samples with 13C-labeled residues at selected positions, namely, V18, F20, A30, I31, G33, L34, and M35. These residues belonged to the b-sheet segments of mature Ab40 fibrils.[17] For comparison, we also prepared Ab40 fibrils at a peptide concentration of 100 mm. Figure 5 shows the overlay spectra of the 13C–13C NMR correlation spectra of the lipoAb-7d and fibril samples. The full NMR spectra and their assignments are given in the Supporting Information (Figures S5–S7 and Table S1). Interestingly, only a single set of correlation signals was observed for lipoAb-7d and the line widths were in the range of 1.2–3.2 ppm. The backbone chemical shifts were very similar to those of the fibril sample (Figure 6). Yet, the 13C NMR chemical shifts of the side chains of V18, A30, L34, and M35 were somewhat different between the two samples. To verify whether the NMR signals of the lipoAb-7d spectrum arose from the majority of the peptides, we used the fibril sample as the NMR spin-counting reference. By comparing the cross-peak intensities (CO¢Ca) of the lipoAb-7d and fibril spectra, we
found that the signal intensity of lipoAb-7d accounted for about 67 % of the peptides (see the Supporting Information). The signal of the remaining peptides, which may exist in the form of monomers and/or unstructured aggregates, is presumably too broad to be detected. Previously, Ishii and co-workers reported that the b-sheet intermediates of Ab40 were structurally similar to the mature fibrils.[18] Therefore, we suggest that the Ab40 peptides of lipoAb-7d are largely folded into the intermediate state that has well-defined fibril-like structures. The difference in the side-chain chemical shifts between the lipoAb and fibril samples may be due to the different packing of the side chains. Nevertheless, we cannot rule out the possibility that the span of the b-strand region of the Ab40 peptides of lipoAb-7d may be slightly different from that of mature fibrils.[19] Do Liposomes Necessarily Affect the Aggregation State of Ab Peptides?
When the Ab peptide solution is mixed with liposomes with anionic lipid head groups, the membrane surface can facilitate the nucleation and growth processes of fibrils.[7–9] Recently, it was demonstrated that liposomes formed by DMPC/DMPG could significantly influence the fibrillization kinetics and molecular structure of Ab40 fibrils, for which the peptides at a concentration of 50 mm were incubated with lipid vesicles of 1.25 mm for 2 h.[14] Indeed, the corresponding chemical shift difference between their fibril and lipid–fibril samples is larger than 4 ppm for some residues (Figure S8 in the Supporting Information). Because of the intimate interaction between the peptide and lipid at the nucleation stage, the molecular structure of the fibrils formed will be modulated substantially by the templating effect of anionic liposomes.[14] In our study, on the other hand, the liposomes formed by zwitterionic lipids simply provided a temporary confinement of the Ab40 peptides. After purification by SEC, the Ab40 peptides of lipoAb were either encapsulated within the liposomes or confined to the interliposomal space. Under these circumstances, the Ab40 peptides could still fold into an intermediate state with a similar molecular structure to the fibrils formed in free solution at a concentration of 100 mm. This 13 13 Figure 5. Overlay of 2D solid-state C– C NMR spectra of Ab40 fibrils (dark grey) and lipoAb-7d (light grey). The observation is remarkable. First, labels indicate assignments on the fibril spectrum. The spectra were plotted with Sparky software, with the contour levels increasing by a factor of 1.4 successively. it demonstrates that the effect Chem. Asian J. 2015, 10, 1967 – 1971
www.chemasianj.org
1969
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper lipid composition. We believe that a thorough understanding of the fibrillization pathway in lipoAb can shed more light on the mechanism of in vivo aggregation of Ab peptides.
Experimental Section
Figure 6. Difference in the 13C chemical shifts between lipoAb-7d and Ab fibrils.
of zwitterionic lipids on the nucleation process of the Ab40 aggregates should be limited, that is, zwitterionic lipids should not induce any significant conformational difference in Ab40 fibrils. Because Ab peptides confined to the intercellular space diffuse more slowly than those in free solution,[24] and the limited amount of peptides would preclude the formation of fibrils, it is likely for the peptide aggregates to adopt a uniform structure. That is, the conformational energy landscape of the peptide aggregates in lipoAb might have a well-defined minimum. Second, our data suggested that, at a bulk concentration lower than the critical micellar concentration (CMC), peptide aggregation could still occur provided that the local concentration was sufficiently high. This may explain the puzzling observations that the CMC of Ab40 has been reported in the range of 0.05–17.5 mm,[20–22] and yet the Ab concentration in the interstitial fluid is in the low nanomolar range even under AD conditions.[23] Thus, it is a reasonable hypothesis that Ab aggregation in vivo is triggered by being concentrated within the intercellular space through membrane-mediated interactions. Furthermore, the Ab40 intermediates formed in lipoAb were as structurally well ordered as mature fibrils. This observation partially justifies attempts to develop drug inhibitors based on the structural models developed from in vitro studies, in which the fibrils were usually incubated in free solution. Further studies are required to verify whether the same phenomenon is observed for Ab42 peptides, which are also closely associated with AD.[10]
Conclusion We developed a liposome–Ab system to probe the aggregation state of Ab40 peptides in close association with zwitterionic lipid bilayer membranes. Although biological membranes are likely to play an important role in the fibrillization process of Ab peptides in vivo, additional studies are required to clarify whether the membranes can affect the molecular structure of the fibrils or merely provide a confined space for peptide aggregation. Interesting issues that warrant further investigation include whether the fibril growth mechanism in the sample of lipoAb-8d-lysis-7d is dominated by a simple coalescence of the peptides folded in the intermediate state, and whether different polymorphs of Ab fibrils can be obtained by varying the Chem. Asian J. 2015, 10, 1967 – 1971
www.chemasianj.org
A detailed description of sample preparation, the incubation procedure, and sample characterization data are given in the Supporting Information. Briefly, a 150 mm solution of Ab40 monomer (2.0 mL) was added to a solution of lipids (6.0 mL; 11 mm in diethyl ether). After drying by rotary evaporation at 35 8C, 20 mm Tris-HCl buffer (1 mL) was added gently. The lipoAb sample was purified by SEC. All NMR spectroscopy experiments were carried out at 13C and 1H frequencies of 201.2 and 800.2 MHz, respectively, on a Bruker Avance III spectrometer equipped with a commercial 3.2 mm Efree probe. 13C NMR chemical shifts were externally referenced to tetramethylsilane (TMS) by using adamantane as a secondary reference. The spinning frequency at the magic angle was 10 kHz and the sample temperature was maintained at 268 K.
Acknowledgements This work was financially supported by the Ministry of Science and Technology (100-2628M-002-009-MY3, 100-2731M-002002-MY2). The NMR spectroscopy, cryo-TEM, and TEM measurements were carried out at the Instrumentation Center of National Taiwan University, supported by the Ministry of Science and Technology. We thank Ching-Yen Lin, Ya-Yun Yang, and C.Y. Chien for their assistance in TEM experiments. Keywords: aggregation · Alzheimer’s disease · amyloid-beta peptides · liposomes · oligomers [1] C. M. Dobson, Semin. Cell Dev. Biol. 2004, 15, 3 – 16. [2] F. Chiti, C. M. Dobson, Annu. Rev. Biochem. 2006, 75, 333 – 366. [3] T. P. J. Knowles, M. Vendruscolo, C. M. Dobson, Nat. Rev. Mol. Cell Biol. 2014, 15, 384 – 396. [4] R. Tycko, Annu. Rev. Phys. Chem. 2011, 62, 279 – 299. [5] R. Tycko, R. B. Wickner, Acc. Chem. Res. 2013, 46, 1487 – 1496. [6] G. P. Gorbenko, P. K. J. Kinnunen, Chem. Phys. Lipids 2006, 141, 72 – 82. [7] K. Matsuzaki, Biochim. Biophys. Acta Biomembr. 2007, 1768, 1935 – 1942. [8] C. Aisenbrey, T. Borowik, R. Bystrçm, M. Bokvist, F. Lindstrçm, H. Misiak, M.-A. Sani, G. Grçbner, Eur. Biophys. J. 2008, 37, 247 – 255. [9] S. M. Butterfield, H. A. Lashuel, Angew. Chem. Int. Ed. 2010, 49, 5628 – 5654; Angew. Chem. 2010, 122, 5760 – 5788. [10] S. A. Kotler, P. Walsh, J. R. Brender, A. Ramamoorthy, Chem. Soc. Rev. 2014, 43, 6692 – 6700. [11] J. McLaurin, A. Chakrabartty, J. Biol. Chem. 1996, 271, 26482 – 26489. [12] L.-P. Choo-Smith, W. K. Surewicz, FEBS Lett. 1997, 402, 95 – 98. [13] R. O. Calderon, B. Attema, G. H. DeVries, J. Neurochem. 1995, 64, 424 – 429. [14] Z. Niu, W. Zhao, Z. Zhang, F. Xiao, X. Tang, J. Yang, Angew. Chem. Int. Ed. 2014, 53, 9294 – 9297. [15] F. Szoka, D. Papahadjopoulos, Proc. Natl. Acad. Sci. USA 1978, 75, 4194 – 4198. [16] K. S. Kim, D. L. Miller, V. J. Sapienza, C. M. J. Chen, C. Bai, I. GrundkeIqbal, J. R. Currie, H. M. Wisniewski, Neurosci. Res. Commun. 1988, 2, 121 – 130. [17] A. K. Paravastua, R. D. Leapman, W. M. Yau, R. Tycko, Proc. Natl. Acad. Sci. USA 2008, 105, 18349 – 18354. [18] S. Chimon, M. A. Shaibat, C. R. Jones, D. C. Calero, B. Aizezi, Y. Ishii, Nat. Struct. Mol. Biol. 2007, 14, 1157 – 1164.
1970
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper [19] H. A. Scheidt, I. Morgado, S. Rothemund, D. Huster, M. Fandrich, Angew. Chem. Int. Ed. 2011, 50, 2837 – 2840. [20] P. Sengupta, K. Garai, B. Sahoo, Y. Shi, D. J. E. Callaway, S. Maiti, Biochemistry 2003, 42, 10506 – 10513. [21] R. Sabat¦, J. Estelrich, J. Phys. Chem. B 2005, 109, 11027 – 11032. [22] W. Qiang, K. Kelley, R. Tycko, J. Am. Chem. Soc. 2013, 135, 6860 – 6871. [23] I. W. Hamley, Chem. Rev. 2012, 112, 5147 – 5192.
Chem. Asian J. 2015, 10, 1967 – 1971
www.chemasianj.org
[24] D. A. Rusakov, D. M. Kullmann, Proc. Natl. Acad. Sci. USA 1998, 95, 8975 – 8980. Manuscript received: May 10, 2015 Revised: June 1, 2015 Accepted Article published: June 10, 2015 Final Article published: July 16, 2015
1971
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper
Structure Determination
Aggregation of Beta-Amyloid Peptides Proximal to Zwitterionic Lipid Bilayers Chien-I Yang,[a] Brook N. F. Tsai,[a] Shing-Jong Huang,[b] Ting-Yu Wang,[b] Hwan-Ching Tai,*[a] and Jerry C. C. Chan*[a] Abstract: One of the hallmarks of Alzheimer’s disease is the deposition of amyloid plaques, which consist of b-amyloid (Ab) peptides in fibrillar states. Nonfibrillar Ab aggregates have been considered as an important intermediate in the pathway of fibrillization, but little is known about the formation mechanism. The on-pathway b-sheet intermediates of Ab40 peptides can be trapped by incubating the peptides in liposomes formed by zwitterionic lipids. The aggregates of Ab40 peptides have been prepared at a peptide concentra-
Introduction Aggregation of b-amyloid (Ab) peptides has been closely associated with the progression of Alzheimer’s disease (AD).[1–3] On the basis of in vitro studies, the molecular structures of Ab40 fibrils have been largely clarified.[4, 5] In the crowded environment of both intra- and extracellular space, both are full of membrane-enclosed organelles and vesicles. It is unlikely for Ab to aggregate in vivo following the type of kinetics it exhibits in free solution. Interactions between Ab peptides and lipid membranes have been studied extensively to unravel the molecular mechanism underlying the pathogenesis of AD.[6–10] Many studies show that membranes containing anionic lipids strongly affect the molecular structure of Ab fibrils.[7, 10] In particular, it has been demonstrated that liposomes formed by dimyristoylphosphatidylcholine (DMPC)/dimyristoylphosphatidylglycerol (DMPG) can significantly influence the fibrillization kinetics and molecular structure of Ab40 fibrils.[14] On the other hand, freshly solubilized Ab does not interact strongly with phosphatidylcholine (PC) at neutral pH.[7, 11, 12] Furthermore, the major phospholipids in neuronal soma are PC and phosphatidylethanolamine (PE).[13] Therefore, it is of interest to study the [a] C.-I. Yang, B. N. F. Tsai, Prof. Dr. H.-C. Tai, Prof. Dr. J. C. C. Chan Department of Chemistry, National Taiwan University No. 1, Section 4, Roosevelt Road, Taipei 106 (Taiwan) E-mail: [email protected] [email protected] [b] Dr. S.-J. Huang, Dr. T.-Y. Wang Instrumentation Center, National Taiwan University No. 1, Section 4, Roosevelt Road, Taipei 106 (Taiwan) Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/asia.201500482. Chem. Asian J. 2015, 10, 1967 – 1971
tion of less than 10 mm. Solid-state NMR spectroscopy data show that the backbone conformation of the aggregates is almost identical to that of the fibrils formed in free solution. In contrast to anionic lipids, zwitterionic lipids, which are typical of neuronal soma, did not induce any significant conformational difference in Ab40 fibrils. This liposome–Ab system may serve as a useful model to study the fibril formation mechanism.
molecular structure of Ab peptides incubated with neutral liposomes. Herein, we show that the 40-residue b-amyloid peptides (Ab40) incubated in the presence of PC/PE liposomes fold into an intermediate state that has a similar structure to that observed for fibrils incubated in free solution.
Results and Discussion Determination of the Peptide Concentration of LipoAb The approach of reverse-phase evaporation[15] was adopted for the preparation of our liposome–peptide sample, namely, lipoAb, from a 150 mm solution of Ab40 monomer. The peptides neither encapsulated by, nor interacting strongly with, liposomes were removed by size exclusion chromatography (SEC), in which two fractions containing liposomes and peptide monomers were well separated (Figure 1). The integrity of the liposomes was monitored by dynamic light scattering (DLS) measurements. The average diameter was found to be about 250 nm with a polydispersity index (PdI) equal to 0.19 0.02 (Figure S1 in the Supporting Information). The amount of Ab40 peptides in lipoAb was quantified with the dot-blot assay by using the 6E10 antibody.[16] A stock solution of Ab40 peptides (150 mm, incubated for 2 days) was prepared and then diluted with 20 mm Tris buffer to 2, 4, 6, 10, and 20 mm. These standard solutions were used to quantify the amount of Ab40 peptides in lipoAb. From the integrals of the fluorescence signals of the standard solutions (Figure 2), the Ab40 concentration of lipoAb was determined to be (9.1 0.3) mm and that of the monomer fraction was (23.2 0.6) mm, whereas the local concentration of Ab40 within a liposome should be close to 150 mm. On the basis of the amount of lipid and the average
1967
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper
Figure 3. TEM image of the lipoAb-1d sample stained with 2 % uranyl acetate. Intact liposomes were readily observed.
Figure 1. Typical SEC profile for sample purification. The purified lipoAb was collected in the fractions of 37.5–57.0 mL. The free Ab40 monomers were eluted at 72.0–82.5 mL.
Figure 2. Signal integrals of the fluorescence intensities of the dot-blot assay of the lipoAb sample, the two-fold diluted monomer fraction, and standard solutions of Ab40.
mained unruptured was consistent with the notion that zwitterionic lipids had relatively weak interactions with Ab peptides. To characterize the aggregation state of the Ab40 peptides by the thioflavin-T (ThT) assay, the lipoAb samples were treated with 1 % (v/v) Triton X-100 to lyse the liposomes. The mature fibrils of each sample, if any, were collected by centrifugation for subsequent ThT measurements. Figure 4 shows the ThT data of lipoAb samples incubated for different periods. The lipoTris sample, which served as a control, was prepared in the absence of Ab40 peptides. As expected, the ThT intensity of lipoAb-1d would be only marginally higher than that of lipoTris. The ThT intensity increased gradually as the incubation period increased up to 15 days. As another control, the lysed lipoAb-8d sample was incubated for an additional 7 days (lipoAb-8d-lysis-7d). The significant enhancement of the ThT intensity provided a strong indication of fibril formation. The corresponding TEM image also confirmed the formation of mature fibrils (Figure S4 in the Supporting Information). Consequently, we concluded that the majority of the peptides in lipoAb were not in a fibrillar state. Indeed, fibrils were sparsely found in the TEM images of the lipoAb samples. The ThT data
diameter of each liposome, we estimated that at least 26 % of the Ab40 peptides were residing outside the liposomes for lipoAb (see the Supporting Information). Characterization of LipoAb The lipoAb samples were incubated at 37 8C under quiescent conditions for various periods from 2 h to 15 days. The incubated samples are henceforth referred to as lipoAb-2 h, and so on. Figure 3 shows the negative-stain TEM image of the lipoAb-1d sample, in which liposomes were readily found, although a minute amount of fibrils were also observed. As expected, the liposomes were multivesicular, as observed in the cryo-TEM image (Figure S2 in the Supporting Information). DLS measurements confirmed that the liposomes were intact during the whole course of the experiment (Figure S3 in the Supporting Information). The fact that most liposomes reChem. Asian J. 2015, 10, 1967 – 1971
www.chemasianj.org
Figure 4. ThT data of lipoAb samples after lysing by 1 % (v/v) Triton.
1968
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper alone, however, could not elucidate whether the peptides were in monomeric form or in certain intermediate states along the fibrillization pathway. Solid-State NMR Spectroscopy Measurements of LipoAb To probe the molecular structure of the peptides in lipoAb, we prepared samples with 13C-labeled residues at selected positions, namely, V18, F20, A30, I31, G33, L34, and M35. These residues belonged to the b-sheet segments of mature Ab40 fibrils.[17] For comparison, we also prepared Ab40 fibrils at a peptide concentration of 100 mm. Figure 5 shows the overlay spectra of the 13C–13C NMR correlation spectra of the lipoAb-7d and fibril samples. The full NMR spectra and their assignments are given in the Supporting Information (Figures S5–S7 and Table S1). Interestingly, only a single set of correlation signals was observed for lipoAb-7d and the line widths were in the range of 1.2–3.2 ppm. The backbone chemical shifts were very similar to those of the fibril sample (Figure 6). Yet, the 13C NMR chemical shifts of the side chains of V18, A30, L34, and M35 were somewhat different between the two samples. To verify whether the NMR signals of the lipoAb-7d spectrum arose from the majority of the peptides, we used the fibril sample as the NMR spin-counting reference. By comparing the cross-peak intensities (CO¢Ca) of the lipoAb-7d and fibril spectra, we
found that the signal intensity of lipoAb-7d accounted for about 67 % of the peptides (see the Supporting Information). The signal of the remaining peptides, which may exist in the form of monomers and/or unstructured aggregates, is presumably too broad to be detected. Previously, Ishii and co-workers reported that the b-sheet intermediates of Ab40 were structurally similar to the mature fibrils.[18] Therefore, we suggest that the Ab40 peptides of lipoAb-7d are largely folded into the intermediate state that has well-defined fibril-like structures. The difference in the side-chain chemical shifts between the lipoAb and fibril samples may be due to the different packing of the side chains. Nevertheless, we cannot rule out the possibility that the span of the b-strand region of the Ab40 peptides of lipoAb-7d may be slightly different from that of mature fibrils.[19] Do Liposomes Necessarily Affect the Aggregation State of Ab Peptides?
When the Ab peptide solution is mixed with liposomes with anionic lipid head groups, the membrane surface can facilitate the nucleation and growth processes of fibrils.[7–9] Recently, it was demonstrated that liposomes formed by DMPC/DMPG could significantly influence the fibrillization kinetics and molecular structure of Ab40 fibrils, for which the peptides at a concentration of 50 mm were incubated with lipid vesicles of 1.25 mm for 2 h.[14] Indeed, the corresponding chemical shift difference between their fibril and lipid–fibril samples is larger than 4 ppm for some residues (Figure S8 in the Supporting Information). Because of the intimate interaction between the peptide and lipid at the nucleation stage, the molecular structure of the fibrils formed will be modulated substantially by the templating effect of anionic liposomes.[14] In our study, on the other hand, the liposomes formed by zwitterionic lipids simply provided a temporary confinement of the Ab40 peptides. After purification by SEC, the Ab40 peptides of lipoAb were either encapsulated within the liposomes or confined to the interliposomal space. Under these circumstances, the Ab40 peptides could still fold into an intermediate state with a similar molecular structure to the fibrils formed in free solution at a concentration of 100 mm. This 13 13 Figure 5. Overlay of 2D solid-state C– C NMR spectra of Ab40 fibrils (dark grey) and lipoAb-7d (light grey). The observation is remarkable. First, labels indicate assignments on the fibril spectrum. The spectra were plotted with Sparky software, with the contour levels increasing by a factor of 1.4 successively. it demonstrates that the effect Chem. Asian J. 2015, 10, 1967 – 1971
www.chemasianj.org
1969
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper lipid composition. We believe that a thorough understanding of the fibrillization pathway in lipoAb can shed more light on the mechanism of in vivo aggregation of Ab peptides.
Experimental Section
Figure 6. Difference in the 13C chemical shifts between lipoAb-7d and Ab fibrils.
of zwitterionic lipids on the nucleation process of the Ab40 aggregates should be limited, that is, zwitterionic lipids should not induce any significant conformational difference in Ab40 fibrils. Because Ab peptides confined to the intercellular space diffuse more slowly than those in free solution,[24] and the limited amount of peptides would preclude the formation of fibrils, it is likely for the peptide aggregates to adopt a uniform structure. That is, the conformational energy landscape of the peptide aggregates in lipoAb might have a well-defined minimum. Second, our data suggested that, at a bulk concentration lower than the critical micellar concentration (CMC), peptide aggregation could still occur provided that the local concentration was sufficiently high. This may explain the puzzling observations that the CMC of Ab40 has been reported in the range of 0.05–17.5 mm,[20–22] and yet the Ab concentration in the interstitial fluid is in the low nanomolar range even under AD conditions.[23] Thus, it is a reasonable hypothesis that Ab aggregation in vivo is triggered by being concentrated within the intercellular space through membrane-mediated interactions. Furthermore, the Ab40 intermediates formed in lipoAb were as structurally well ordered as mature fibrils. This observation partially justifies attempts to develop drug inhibitors based on the structural models developed from in vitro studies, in which the fibrils were usually incubated in free solution. Further studies are required to verify whether the same phenomenon is observed for Ab42 peptides, which are also closely associated with AD.[10]
Conclusion We developed a liposome–Ab system to probe the aggregation state of Ab40 peptides in close association with zwitterionic lipid bilayer membranes. Although biological membranes are likely to play an important role in the fibrillization process of Ab peptides in vivo, additional studies are required to clarify whether the membranes can affect the molecular structure of the fibrils or merely provide a confined space for peptide aggregation. Interesting issues that warrant further investigation include whether the fibril growth mechanism in the sample of lipoAb-8d-lysis-7d is dominated by a simple coalescence of the peptides folded in the intermediate state, and whether different polymorphs of Ab fibrils can be obtained by varying the Chem. Asian J. 2015, 10, 1967 – 1971
www.chemasianj.org
A detailed description of sample preparation, the incubation procedure, and sample characterization data are given in the Supporting Information. Briefly, a 150 mm solution of Ab40 monomer (2.0 mL) was added to a solution of lipids (6.0 mL; 11 mm in diethyl ether). After drying by rotary evaporation at 35 8C, 20 mm Tris-HCl buffer (1 mL) was added gently. The lipoAb sample was purified by SEC. All NMR spectroscopy experiments were carried out at 13C and 1H frequencies of 201.2 and 800.2 MHz, respectively, on a Bruker Avance III spectrometer equipped with a commercial 3.2 mm Efree probe. 13C NMR chemical shifts were externally referenced to tetramethylsilane (TMS) by using adamantane as a secondary reference. The spinning frequency at the magic angle was 10 kHz and the sample temperature was maintained at 268 K.
Acknowledgements This work was financially supported by the Ministry of Science and Technology (100-2628M-002-009-MY3, 100-2731M-002002-MY2). The NMR spectroscopy, cryo-TEM, and TEM measurements were carried out at the Instrumentation Center of National Taiwan University, supported by the Ministry of Science and Technology. We thank Ching-Yen Lin, Ya-Yun Yang, and C.Y. Chien for their assistance in TEM experiments. Keywords: aggregation · Alzheimer’s disease · amyloid-beta peptides · liposomes · oligomers [1] C. M. Dobson, Semin. Cell Dev. Biol. 2004, 15, 3 – 16. [2] F. Chiti, C. M. Dobson, Annu. Rev. Biochem. 2006, 75, 333 – 366. [3] T. P. J. Knowles, M. Vendruscolo, C. M. Dobson, Nat. Rev. Mol. Cell Biol. 2014, 15, 384 – 396. [4] R. Tycko, Annu. Rev. Phys. Chem. 2011, 62, 279 – 299. [5] R. Tycko, R. B. Wickner, Acc. Chem. Res. 2013, 46, 1487 – 1496. [6] G. P. Gorbenko, P. K. J. Kinnunen, Chem. Phys. Lipids 2006, 141, 72 – 82. [7] K. Matsuzaki, Biochim. Biophys. Acta Biomembr. 2007, 1768, 1935 – 1942. [8] C. Aisenbrey, T. Borowik, R. Bystrçm, M. Bokvist, F. Lindstrçm, H. Misiak, M.-A. Sani, G. Grçbner, Eur. Biophys. J. 2008, 37, 247 – 255. [9] S. M. Butterfield, H. A. Lashuel, Angew. Chem. Int. Ed. 2010, 49, 5628 – 5654; Angew. Chem. 2010, 122, 5760 – 5788. [10] S. A. Kotler, P. Walsh, J. R. Brender, A. Ramamoorthy, Chem. Soc. Rev. 2014, 43, 6692 – 6700. [11] J. McLaurin, A. Chakrabartty, J. Biol. Chem. 1996, 271, 26482 – 26489. [12] L.-P. Choo-Smith, W. K. Surewicz, FEBS Lett. 1997, 402, 95 – 98. [13] R. O. Calderon, B. Attema, G. H. DeVries, J. Neurochem. 1995, 64, 424 – 429. [14] Z. Niu, W. Zhao, Z. Zhang, F. Xiao, X. Tang, J. Yang, Angew. Chem. Int. Ed. 2014, 53, 9294 – 9297. [15] F. Szoka, D. Papahadjopoulos, Proc. Natl. Acad. Sci. USA 1978, 75, 4194 – 4198. [16] K. S. Kim, D. L. Miller, V. J. Sapienza, C. M. J. Chen, C. Bai, I. GrundkeIqbal, J. R. Currie, H. M. Wisniewski, Neurosci. Res. Commun. 1988, 2, 121 – 130. [17] A. K. Paravastua, R. D. Leapman, W. M. Yau, R. Tycko, Proc. Natl. Acad. Sci. USA 2008, 105, 18349 – 18354. [18] S. Chimon, M. A. Shaibat, C. R. Jones, D. C. Calero, B. Aizezi, Y. Ishii, Nat. Struct. Mol. Biol. 2007, 14, 1157 – 1164.
1970
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper [19] H. A. Scheidt, I. Morgado, S. Rothemund, D. Huster, M. Fandrich, Angew. Chem. Int. Ed. 2011, 50, 2837 – 2840. [20] P. Sengupta, K. Garai, B. Sahoo, Y. Shi, D. J. E. Callaway, S. Maiti, Biochemistry 2003, 42, 10506 – 10513. [21] R. Sabat¦, J. Estelrich, J. Phys. Chem. B 2005, 109, 11027 – 11032. [22] W. Qiang, K. Kelley, R. Tycko, J. Am. Chem. Soc. 2013, 135, 6860 – 6871. [23] I. W. Hamley, Chem. Rev. 2012, 112, 5147 – 5192.
Chem. Asian J. 2015, 10, 1967 – 1971
www.chemasianj.org
[24] D. A. Rusakov, D. M. Kullmann, Proc. Natl. Acad. Sci. USA 1998, 95, 8975 – 8980. Manuscript received: May 10, 2015 Revised: June 1, 2015 Accepted Article published: June 10, 2015 Final Article published: July 16, 2015
1971
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
