Báo cáo khoa học: Lipid-induced conformational transition of the amyloid core fragment Ab(28–35) and its A30G and A30I mutants docx

Lipid-induced conformational transition of the amyloid core fragment Ab(28–35) and its A30G and A30I mutants Sureshbabu Nagarajan 1 , Kirubagaran Ramalingam 2 , P. Neelakanta Reddy 1 , Damiano M. Cereghetti 3 , E. J. Padma Malar 4 and Jayakumar Rajadas 1 1 Bio-Organic and Neurochemistry Laboratory, Central Leather Research Institute, Adyar, Chennai, India 2 National Institute of Ocean Technology, Pallikaranai, Chennai, India 3 Department of Medicine, Stanford University, CA, USA 4 National Centre for Ultrafast Processes, University of Madras, Chennai, India Keywords amyloid core fragment Ab(28–35); hydrophobicity and sheet propensity; membrane disruption and neurotoxicity; mutation; negatively charged lipids Correspondence J. Rajadas, Bio-Organic and Neurochemistry Laboratory, Central Leather Research Institute, Adyar, Chennai 600 020 Fax: +91 44 24911589 Tel: +91 44 24911386 extn 324 E-mail: karkuvi77@yahoo.co.uk (Received 9 November 2007, revised 8 February 2008, accepted 5 March 2008) doi:10.1111/j.1742-4658.2008.06378.x The interaction of the b-amyloid peptide (Ab) with neuronal membranes could play a key role in the pathogenesis of Alzheimer’s disease. Recent studies have focused on the interactions of Ab oligomers to explain the neuronal toxicity accompanying Alzheimer’s disease. In our study, we have investigated the role of lipid interactions with soluble Ab(28–35) (wild-type) and its mutants A30G and A30I in their aggregation and con- formational preferences. CD and Trp fluorescence spectroscopic studies indicated that, immediately on dissolution, these peptides adopted a ran- dom coil structure. Upon addition of negatively charged 1,2-dipalmitoyl- syn-glycero-3-phospho-rac-(glycerol) sodium salt (PG) lipid, the wild-type and A30I mutant underwent reorganization into a predominant b-sheet structure. However, no conformational changes were observed in the A30G mutant on interaction with PG. In contrast, the presence of zwit- terionic 1,2-dipalmitoyl-syn-glycero-3-phosphatidylcholine (PC) lipid had no effect on the conformation of these three peptides. These observations were also confirmed with atomic force microscopy and the thioflavin-T assay. In the presence of PG vesicles, both the wild-type and A30I mutant formed fibrillar structures within 2 days of incubation in NaCl ⁄ P i , but not in their absence. Again, no oligomerization was observed with PC vesi- cles. The Trp studies also revealed that both ends of the three peptides are not buried deep in the vesicle membrane. Furthermore, fluorescence spectroscopy using the environment-sensitive probe 1,6-diphenyl-1,3,5-hex- atriene showed an increase in the membrane fluidity upon exposure of the vesicles to the peptides. The latter effect may result from the lipid head group interactions with the peptides. Fluorescence resonance energy trans- fer experiments revealed that these peptides undergo a random coil-to- sheet conversion in solution on aging and that this process is accelerated by negatively charged lipid vesicles. These results indicate that aggregation depends on hydrophobicity and propensity to form b-sheets of the amy- loid peptide, and thus offer new insights into the mechanism of amyloid neurodegenerative disease. Abbreviations AFM, atomic force microscopy; Ab, b-amyloid peptide; DPH, 1,6-diphenyl-1,3,5-hexatriene; FRET, fluorescence resonance energy transfer; PC, 1,2-dipalmitoyl-syn-glycero-3-phosphatidylcholine; PG, 1,2-dipalmitoyl-syn-glycero-3-phospho-rac-(glycerol) sodium salt; PrP, prion protein; Tht, thioflavin-T. FEBS Journal 275 (2008) 2415–2427 ª 2008 The Authors Journal compilation ª 2008 FEBS 2415 One of the neuropathological features in the brain of patients with Alzheimer’s disease is the presence of extracellular amyloid plaques that are primarily com- posed of a 39–43 residue peptide known as the b-amy- loid peptide (Ab) [1–3]. Intermixed with this are also shorter fragments of peptides [4]. Administration of Ab and its fragments to cultured cells and living tis- sues damages their functionality and compromises their viability [5–7]. The b-amyloid peptides are known to interact strongly with the lipid bilayer [8] as well as metal cations [9,10], thereby possibly initiating cyto- toxic events. The conversion of soluble b-amyloid pep- tides into amyloid fibrils in vitro has been shown to occur via a nucleation-dependent mechanism [11,12]. However, another pathway is likely to be followed in the presence of lipids [13,14]. Many misfolded proteins that are produced during normal protein processing become capable of inducing cytotoxic effects via interaction with cytosolic mem- branes [15–18]. Such cytotoxic proteins tend to contain a significant number of exposed hydrophobic residues, and are often classed as hydrophobic or amphipathic proteins [19]. In addition to their hydrophobic nature, these peptides are often positively charged, and this enables them to interact with negatively charged lipid membranes [20]. We have chosen a key hydrophobic region of Ab that extends over residues 28–35 (KGAIIGLM) [Ab(28–35)] and displays amphipathic properties. This region is thought to play an important role in determining the secondary structure and the neurotoxicity of the protein [21]. Using H ⁄ D exchange NMR spectroscopy, Ippel et al. have shown that this fragment forms a rigid amyloid core [22]. New insights into the components that mediate the self-assembly of various polypeptides into amyloid fibrils will help to answer questions of medical as well as technological interest. Thus, much attention has been devoted to the study of minimal amyloid-forming fragments [23]. As short peptides are easy to design and synthesize, they serve as an excellent model system for studying amyloid fibril formation in particular and biological self-assembly processes in general. It has been shown that mutations within Ab(25–35) reduce the b-sheet content and fibrillogenic properties of the peptide [24,25]. Pike et al. showed that Ab(28–35) modulated both secondary structure and neurotoxicity. Hence, we decided to investigate the effects of substi- tuting Gly and Ile for Ala on the aggregation behavior of the peptide, both in the presence and in the absence of charged and zwitterionic lipids. Substantial evidence has been provided suggesting that electrostatic interactions between the positively charged residues of Ab and the negatively charged membranes might be responsible for the toxic effect of the former on neuronal cells [26,27]. However, there is also provisional evidence that, due to their hydropho- bicity, the amino acids are most likely embedded in the membrane, thereby causing membrane destabiliza- tion and leakage [28–30]. In order to address the rela- tionship between hydrophobicity and propensity to form b-sheets in the presence of biological membranes, we investigated the structure of Ab(28–35) and its mutants A30G and A30I, using various biophysical techniques. Hence, our study was focused on the con- formational transitions occurring in A b(28–35) as a function of both the lipid nature (either negatively charged or zwitterionic) and the degree of hydropho- bicity of the peptide. We employed CD, Trp fluores- cence and acryl amide quenching to evaluate the peptide interaction with the lipids. These results illus- trate that peptide sequence and membrane composition dramatically influence protein assembly (or misassem- bly) at membrane interfaces. Results Induction of b-sheet conformation by acidic phospholipids and the effect on b-sheet formation of substituting Ile and Gly for Ala The far-UV CD spectra of freshly prepared, soluble wild-type (WT) Ab(28–35) and its mutants A30G and A30I (Table 1) in NaCl ⁄ P i support a random coil con- formation with negative minima around 197 nm and a shoulder peak at 225 nm (Fig. 1A). This shoulder peak results from the minor contributions of b-sheet and b-turn structures. Addition of negatively charged 1,2- dipalmitoyl-syn-glycero-3-phospho-rac-(glycerol) sodium salt (PG) vesicles induced conformational changes in the WT peptide and its mutants. The CD spectra of the WT and A30I peptides showed a b-sheet structure in the presence of PG vesicles. For the A30G mutant, a reduction of the coil peak at 197 nm and a concomi- tant increase of the helix and sheet peak at 222 nm were observed (Fig. 1B). Interestingly, the increase in b-sheet percentage was linearly proportional to the increase in hydrophobicity (Gly < Ala <Ile) (Fig. 1B). It should be noted that the CD spectra of the three peptides are not identical in the aggregated state, suggesting that the three peptides form different types of b-structures. The crossover points from ran- dom coil to b-sheet are also different, with k = 195 nm and h 217 = )9443 for the WT peptide, and k = 208 nm and h 217 = )14 109 for the A30I mutant. The facts that all of these peptides form ran- dom coils in solution and are positively charged, and Interactions of Ab(28–35) and its mutants with lipids S. Nagarajan et al. 2416 FEBS Journal 275 (2008) 2415–2427 ª 2008 The Authors Journal compilation ª 2008 FEBS that they show b-sheet structure in the presence of neg- atively charged lipids, suggest that these peptides are membrane-bound. That this association is not transient is hinted at by the observed changes in the membrane fluidity upon exposure of the vesicles to the peptides [see the 1,6-diphenyl-1,3,5-hexatriene (DPH) studies below]. Confirming the involvement of electrostatic charges in the conformational transitions of the peptides are the observations made with zwitterionic phospholipid vesicles. In this case, no effect on structure was observed, and the peptides adopted a predominantly random coil conformation with negative minima at 197 nm (Fig. 1C). We must note, however, that the possibility exists that the corresponding CD spectra result from a polyproline type II b-turn. Similar experiments were performed with both Trp derivatives of the WT peptide and its two mutants (see Table 1 for the sequences). The resulting CD spectra showed that an additional Trp residue at either termi- nus did not have any significant effect on the physico- chemical properties exhibited by the parent molecules (data not shown). These results therefore excluded the possibility that the Trp-modified peptides used in the subsequent experiments (see below) behaved differently from the WT peptide and its two mutants. Thioflavin-T (ThT) assay The presence of large aggregates can be detected by monitoring the binding of dyes such as Congo red or ThT [31]. These dyes are known to bind specifically to the cross-b-structure in a variety of amyloids, yet they do not bind to monomers. Quantification of the aggre- gation extent was done here by the ThT assay. In agreement with the CD results, the ThT fluorescence intensity increased drastically when the WT peptide was dissolved in the presence of PG vesicles (Fig. 2A). This effect was more pronounced for the A30I mutant (Fig. 2C). In contrast, the A30G ThT peak did not show any significant increase when compared to the signal generated by PG vesicles alone (Fig. 2B). Freshly prepared, soluble WT, A30G and A30I pep- tides, either alone or in the presence of 1,2-dipalmi- toyl-syn-glycero-3-phosphatidylcholine (PC) vesicles, did not have any significant effect on the ThT fluores- cence (Fig. 2D–F). Atomic force microscopy (AFM) studies To determine whether PG vesicles promote assembly of the WT, A30G and A30I peptides into amyloid fibers, we employed AFM to examine the structural patterns obtained upon exposure of these three pep- tides to lipids (Fig. 3). In the presence of PC vesicles, the WT, A30G and A30I peptides did not form any detectable amyloid fibers within 2 days of incubation. 10 000 A B C 190 210 230 Wavelength (nm) Peptides alone Peptides + PG Peptides + PC Wavelength (nm) Wavelength (nm) 250 190 210 230 250 190 210 230 250 40 000 30 000 20 000 10 000 0 20 000 10 000 0 –10 000 [θ] deg cm 2 dmol –1 [θ] deg cm 2 dmol –1 [θ] deg cm 2 dmol –1 –20 000 –10 000 –20 000 –30 000 –10 000 –20 000 –30 000 –40 000 –50 000 0 Fig. 1. CD spectra were acquired for 50 l M WT ( ), A30G ( ) and A30I ( ) peptides in NaCl ⁄ P i (pH 7.4) alone (A) and in the presence of PG (B) or PC (C) at a 1 : 30 peptide ⁄ lipid ratio. Differences in the CD spectra demonstrate that WT, A30G and A30I peptides adopt a random structure in NaCl ⁄ P i alone and in the presence of PC lipid. However, a conformational transition is observable in the presence of the PG lipid. Table 1. Peptides used in this study. No. Peptide Primary sequence Relative molecular mass Theoretical Found 1Ab(28–35) KGAIIGLM 802.0 803.9 1a Ab(28–35) ⁄ W27 WKGAIIGLM 988.3 988.6 1b Ab(28–35) ⁄ W36 KGAIIGLMW 988.3 988.7 1c Dansyl-Ab(28–35) ⁄ W36 Dansyl– KGAIIGLMW 1035.7 1036.9 2Ab(28–35) A30G KGGIIGLM 788.0 789.6 2a Ab(28–35) A30G ⁄ W27 WKGGIIGLM 974.2 974.6 2b Ab(28–35) A30G ⁄ W36 KGGIIGLMW 974.2 974.7 2c Dansyl-Ab(28–35) A30G ⁄ W36 Dansyl– KGGIIGLMW 1022.9 1023.4 3Ab(28–35) A30I KGIIIGLM 844.1 845.5 3a Ab(28–35) A30I ⁄ W27 WKGIIIGLM 1030.3 1031.1 3b Ab(28–35) A30I ⁄ W36 KGIIIGLMW 1030.8 1031.5 3c Dansyl-Ab(28–35) A30I ⁄ W36 Dansyl- KGIIIGLMW 1078.8 1079.6 S. Nagarajan et al. Interactions of Ab(28–35) and its mutants with lipids FEBS Journal 275 (2008) 2415–2427 ª 2008 The Authors Journal compilation ª 2008 FEBS 2417 Instead, globular structures with an average radius of 10–50 nm were observed (Fig. 3A–C,G–I). On the other hand, abundant fibers were detected after 2 days for the WT and A30I peptides in the presence of PG vesicles, with an average extension of 200–500 nm (Fig. 3D,F). As expected from the experiments described above, the AFM images of the A30G mutant did not reveal any fibrils but only amorphous aggregates with an average radius of 150–200 nm (Fig. 3E). Effect of site mutation on Trp fluorescence The characteristic fluorescence emission of Trp is highly sensitive to changes in the environment, and it has therefore been widely used to monitor the interac- tion of proteins with lipid membranes [32]. We studied changes of the intrinsic Trp fluorescence with six pep- tides derived from N-labeling and C-labeling of the original peptides (WT, A30G and A30I peptides; see Table 1). When no lipid vesicles were used and the peptides were mainly in a monomeric state, the Trp emission spectra of both peptide derivatives showed maxima at 364, 362 and 365 nm for the WT ⁄ W27, A30G ⁄ W27 and A30I ⁄ W27 peptides, respectively (Fig. 4A–C), indicating that the Trp was highly solvent-exposed. On the contrary, important spectral changes were observed in the presence of the nega- tively charged lipid vesicles (Fig. 4A–C). These changes were characterized by a decreased fluorescence inten- A WT No lipid 500 nm 500 nm 500 nm 500 nm 500 nm 500 nm 500 nm 500 nm 500 nm PG PC A30G A30l BC DEF GHI Fig. 3. AFM images showing the formation of fibrils and aggre- gates of WT, A30G and A30I peptides in NaCl ⁄ P i alone (A–C) and in the the presence of PG (D–F) and PC (G–I). After 2 days of incu- bation in the absence of lipid vesicles, oligomeric species were visi- ble for all peptides (A–C). Peptide fibrils formed in the presence of PG vesicles (D–F), whereas globular aggregates were visible in the presence of PC vesicles (G–I). 200 180 60 50 40 30 20 10 0 450 500 550 160 140 120 100 80 60 40 20 450 450 20 AB C 18 Tht A30G A30l WT 16 14 12 10 8 6 4 2 0 500 550 Wavelength (nm) Wavelength (nm) Peptides + PC Peptides + PGPeptides in solution Tht fluorescence (a.u.) Tht fluorescence (a.u.) Tht fluorescence (a.u.) Wavelength (nm) 500 PG WT+PG WT+PC A30G+PC A30l+PC PC A30G+PG A30l+PG 550 0 Fig. 2. ThT fluorescence spectra. WT, A30G and A30I peptides were incubated at a concentration of 10 l M with 5 lM ThT for 30 min in NaCl ⁄ P i alone or in the presence of PG (A–C) or PC (D–F). Formation of amyloid fibrils, as shown by increased ThT absorption, is particularly evident in the presence of the PG lipids. 150 150 100 50 0 125 100 75 50 25 Tryptophan fluorescence intensity (a.u.) Wavelength (nm) A B D C WT/W@27 A30l/W@27 W@36 A30G/W@27 Wavelength (nm) Wavelength (nm) Wavelength (nm) Tryptophan fluorescence intensity (a.u.) Tryptophan fluorescence intensity (a.u.) Tryptophan fluorescence intensity (a.u.) 0 300 350 400 450 400 300 140 120 100 80 60 40 20 0 200 100 0 300 350 400 450 300 350 400 450 300 350 400 450 Fig. 4. Trp fluorescence spectra of the N-modified WT ⁄ W27 (A), A30G ⁄ W27 (B) and A30I ⁄ W27 (C) peptides in NaCl ⁄ P i alone ( ) and in the presence of PG vesicles ( ). Data for the C-modified peptides are summarized in (D). The excitation wavelength was set to 280 nm, and the fluorescence emission was monitored between 300 and 500 nm. The PG vesicles cause a blue shift of the fluores- cence emission maximum and either a decrease (A–C) or increase (D) in the fluorescence intensity. Interactions of Ab(28–35) and its mutants with lipids S. Nagarajan et al. 2418 FEBS Journal 275 (2008) 2415–2427 ª 2008 The Authors Journal compilation ª 2008 FEBS sity and a spectral blue shift of the peak maxima of 14, 8 and 19 nm for the WT ⁄ W27, A30G ⁄ W27 and A30I ⁄ W27 peptides, respectively (Fig. 4A–C). On the other hand, the spectra of the C-labeled peptides showed an increase in the Trp fluorescence and a less marked blue shift of about 8 nm (Fig. 4D). The more pronounced blue shift for the N-modified peptides suggests that Trp experiences a less polar environment at the N-terminus than at the C-terminus, and that it is located in an environment with increased microviscosity. The increased fluorescence intensity observed with the Trp36 peptides is probably due to the reduced degrees of freedom that the peptides experience in association with the membrane, thereby leading to increased quantum yields. The fact that we see a diminished signal in the N-labeled series may be due to internal quenching by the Lys next to the Trp [33]. Acrylamide quenching studies A prerequisite for any understanding of the interaction of the peptide with the membrane is the knowledge of its location on the membrane. Hence, the location of Trp in negatively charged lipid vesicles was studied by adding increasing amounts of an acrylamide solution and monitoring the resulting quenching of fluorescence in the absence and presence of lipid vesicles. High Stern–Volmer constant (K sv ) values of 5.71, 5.13 and 5.16 (WT ⁄ W27–WT ⁄ W36), A30G ⁄ W27– A30G ⁄ W36 and A30I ⁄ W27–A30I ⁄ W36, respectively) were obtained for the six N-labeled and C-labeled pep- tides in aqueous solution (Fig. 5). When the N-labeled peptides were incubated with PG lipids, the K sv values decreased to 2.81, 2.74 and 2.63 (WT ⁄ W27, A30G ⁄ W27 and A30I ⁄ W27 peptides, respectively). In the case of the C-labeled probes, the K sv values (3.89, 4.14 and 3.16 for the WT ⁄ W36, A30G ⁄ W36 and A30G ⁄ W36 peptides, respectively) were between those of the former two experiments. These data show that upon interaction with the vesicles, the Trp is shielded from the surrounding aqueous environment and it is not easily reached by the quencher. The differences observed between N-labeled and C-labeled peptides may be ascribed to different factors. For instance, it is possible that the two amino groups in the WT ⁄ W27, A30G ⁄ W27 and A30I ⁄ W27 peptides bring the Trp moiety closer to the solution–lipid interface, where it is less accessible to acrylamide. An effect due to internal quenching in the WT ⁄ W27, A30G ⁄ W27 and A30I ⁄ W27 series, due to the closer proximity of the e- amino and a-amino groups, can be ruled out on the basis of the results obtained in the absence of vesicles. Effect of Ab(28–35) on DPH anisotropy in acidic phospholipids It has been reported that the binding of the amyloid protein of AD to lipid membranes can change their fluidity [34]. Therefore, we examined the effect of bind- ing of the WT, A30G and A30I peptides on the fluid- ity of PG lipid vesicles. The relative fluidity of PG vesicles was considered to be gel-like, as indicated by an r-value close to 0.21. However, the DPH anisotropy constant measured after 1 h of incubation with the WT, A30G and A30I peptides significantly decreased, hence pointing to an enhanced internal fluidity of the bilayer (Fig. 6). Fluorescence resonance energy transfer (FRET) assays FRET has been used as a so-called spectroscopic ruler to monitor self-association and to measure distances within proteins and other macromolecules [35]. In order to understand the folding and unfolding of peptides, both in the presence and the absence of PG lipids, FRET measurements were carried out. As there is no intrinsic fluorophore in Ab(28–35), we chose Trp 2 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1 1.8 AB C y = 5.7143x + 1.0114 y = 5.1286x + 1.0211 y = 4.1429x + 1.0064 y = 5.1571x + 1.0232 y = 3.1686x + 0.9911 y = 2.63x + 0.9762 y = 2.7429x + 09786 y = 3.8857x + 1.0057 y = 2.8143x + 0.9989 WT A30G A30l 1.6 1.4 2 1.8 1.6 1.4 1.2 1 0 0.05 [Acrylamide] (m M) [Acrylamide] (mM) [Acrylamide] (m M) 0.1 0.15 0 0.05 0.1 0.15 0 0.05 0.1 0.15 1.2 1 l 0 /l l 0 /l l 0 /l Fig. 5. Stern–Volmer plots for the acrylamide-mediated quenching of the fluorescence signal in the Trp derivatives of WT (A), A30G (B) and A30I (C) peptides. The fluorescence emission was mea- sured at either about 360 nm (in the absence of PG) or at 348 nm (in the presence of PG). In the absence of PG, both N-modified and C-modified peptides gave similar curves, and only that of the Trp27 peptide (r) is shown. Major differences were observed when the peptides were incubated with PG, as shown by the curves obtained for theTrp36 peptide ( ) and the Trp27 peptide ( ). S. Nagarajan et al. Interactions of Ab(28–35) and its mutants with lipids FEBS Journal 275 (2008) 2415–2427 ª 2008 The Authors Journal compilation ª 2008 FEBS 2419 and dansyl as the donor–acceptor pair. Samples were prepared such that labeled and unlabeled peptides were present in a 1 : 8 ratio to ensure that the FRET was predominantly intramolecular. For peptides in solution, the Trp fluorescence of the dansyl-Ab (28–35) ⁄ W36 derivatives dansyl-WT ⁄ W36, dansyl- A30G ⁄ W36 and dansyl-A30I ⁄ W36 was less than that of the corresponding peptides not conjugated to the dansyl group, WT ⁄ W36, A30G ⁄ W36, and A30G ⁄ W36 (Fig. 7D–F and A–C, respectively). Upon aging, and with no vesicles present, the Trp fluorescence intensity increased, with a concomitant decrease in dansyl fluo- rescence, indicating an increased end-to-end distance. Generally, formation of b-sheet is accompanied by an increase in the intramolecular fluorophore distance. Surprisingly, upon binding to PG vesicles, an increase in energy transfer was also noted, as indicated by a decrease in Trp intensity and an increase in dansyl flu- orescence as compared to peptides in solution (Fig. 8). Only a marginal change in the energy transfer was observed upon prolonged incubation. The very high dansyl ⁄ Trp fluorescence intensity ratio obtained in the latter case questions the formation of a b-sheet. How- ever, the data can be reconciled by considering anti- parallel b-sheet aggregates. In this event, peptides would self-associate along the surface in an alternate way, whereby a peptide bound to the membrane via its two amino groups is flanked by a peptide in the oppo- site direction, and so on. This organization would bring donor and acceptor in close proximity, thus explaining the increased dansyl ⁄ Trp fluorescence inten- sity ratio. This mechanism would be particularly plau- sible with the dansyl derivatives, as the presence of the aromatic fluorophore at the N-terminus removes a basic group and introduces a sulfamoyl group. The latter may lead to a destabilized interaction with the phosphate heads in the bilayer. The tertiary amine present in the fluorophore is probably not ideally posi- tioned and ⁄ or strong enough to efficiently bind to the membrane and therefore counterbalance the desta- bilizing effect. Discussion A prerequisite for a peptide to interact with lipids is the presence of an exposed hydrophobic region that can be stabilized by an amphipathic environment such as a lipid membrane. In this study, we designed and tested the properties of the amyloid core fragment Ab(28–35) and two of its mutants, A30G and A30I, to understand the peptide–membrane interactions, and especially the contributions of the hydrophobic resi- dues Ala, Gly and Ile to this process. Previous studies have demonstrated that acidic phospholipid and phos- phoinositides promote a conformational transition in Ab from a random to a b-structure [36–41]. Figures 1 and 2 show that mutants with a more hydrophobic character have higher propensity to aggregate than the WT sequence in the presence of negatively charged PG vesicles. The greater hydrophobicity of Ile relative to Ala at position 30 presumably accounts for this enhanced aggregation. Likewise, for the A30G mutant, the lower hydrophobicity of Gly may account for the decreased aggregation of this mutant relative to the wild-type. These results indicate that the hydrophobic- ity of the amino acid at position 30 is a major contrib- utor to the enhanced amyloidogenicity. These observations strongly argue that hydrogen bonding within b-structures and hydrophobic interactions between side-chains are likely to be major stabilizing interactions within aggregates. Therefore, increases in the propensities for such interactions are likely to enhance the rate at which aggregation occurs. Hence, these interactions would presumably be very sensitive to the size and character of the hydrophobic residue. The link between structural alteration and membrane destabilization is confirmed by point mutations in the prion protein (PrP) hydrophobic region. The decrease in b-sheet content closely corresponds to decreased cytoxicity. Jobling et al. [42] substituted the hydropho- bic residues Ala and Val with the hydrophilic residue Ser in the PrP(106–126) hydrophobic core (resi- dues 113–122). This substitution induced a reduction in the hydrophobicity of region 113–122 and dramati- cally reduced the neurotoxicity of PrP(106–126). Our investigation provides insights into the role of Trp in peptide aggregation and interactions with lipids. According to Fig. 9, the kinetic changes in Trp 0.17 0.18 0.19 0.2 0.21 0.22 0.23 Steady state anisotropy Control WT A30G A30I Fig. 6. Effect of WT, A30G and A30I peptides on membrane fluid- ity of PG vesicles as determined by DPH anisotropy. The addition of 10 l M peptides to lipid vesicles increases the membrane fluidity. Interactions of Ab(28–35) and its mutants with lipids S. Nagarajan et al. 2420 FEBS Journal 275 (2008) 2415–2427 ª 2008 The Authors Journal compilation ª 2008 FEBS emission for the WT ⁄ W36, A30G ⁄ W36 and A30G ⁄ W36 peptides follow a sigmoidal shape, indicat- ing a cooperative process in solution. However, inter- action with PG lipids decreases the activation barrier via favorable electrostatic and hydrophobic forces. This is clearly seen for the WT and A30I peptides, where the fluorescence intensity increases during the first 24 and 48 h, and then gradually decreases. The less hydrophobic A30G mutant reached a maximum only after 5 days (Fig. 9). Furthermore, the recruit- ment of the peptide to the surface of the PG lipid is rapid, as evidenced by an immediate increase in ThT fluorescence. Terzi et al. have reported on the impor- tance of electrostatic interactions for Ab(25–35) bind- ing to negatively charged liposomes [27]. By forming a b-sheet scaffold structure, Ab can reside on the surface of the lipid head group and self-associate to form the critical fibril nucleus. After nucleation, the fibril grows through the lipid bilayer, ultimately destabilizing the membrane, as indicated by the increased membrane fluidity (Fig. 6). The information on the orientation of the peptide relative to the membrane was acquired by using two probes labeled with Trp at two different positions: the N-terminus and the C-terminus, respec- tively. The differences noticed between the Trp fluores- cence emission blue shifts and the K sv values of the A B C D E F 150 0 h WT/W@36 A30G/W@36 A30l/W@36 Dan– WT/W@36 Dan–A30G/W@36 Dan–A30l/W@36 24 h 48 h 72 h 96 h 120 h 0 h 24 h 48 h 72 h 96 h 120 h 0 h 24 h 48 h 72 h 96 h 120 h 0 h 24 h 48 h 72 h 96 h 120 h 0 h 24 h 48 h 72 h 96 h 120 h 0 h 24 h 48 h 72 h 96 h 120 h 160 140 120 100 80 60 40 20 0 100 100 140 120 200 180 160 140 120 80 70 60 50 40 30 20 10 0 100 80 60 40 20 0 100 80 60 40 20 0 50 0 300 350 400 450 300 400 500 300 350 400 450 300 400 500 50 300 350 Wavelength (nm) Fluorescence intensity (a.u.) Fluorescence intensity (a.u.) Fluorescence intensity (a.u.) Fluorescence intensity (a.u.) Fluorescence intensity (a.u.) Fluorescence intensity (a.u.) Wavelength (nm) Wavelength (nm) Wavelength (nm) Wavelen g th (nm) Wavelen g th (nm) 400 450 300 400 500 0 Fig. 7. Time dependence of FRET for the dansyl-Trp peptide derivatives dansyl- WT ⁄ W36, dansyl-A30G ⁄ W36, and dansyl- A30I ⁄ W36 (D–F), and for their negative controls WT ⁄ W36, A30G ⁄ W36, and A30I ⁄ W36 (A–C). Spectra were acquired at different time intervals (0, 24, 48, 72, 96 and 120 h) by exciting Trp at 280 nm and recording the emission between 300 and 535 nm at 298 K. S. Nagarajan et al. Interactions of Ab(28–35) and its mutants with lipids FEBS Journal 275 (2008) 2415–2427 ª 2008 The Authors Journal compilation ª 2008 FEBS 2421 WT ⁄ W27–WT ⁄ W36, A30G ⁄ W27–A30G ⁄ W36 and A30I ⁄ W27–A30I ⁄ W36 peptides are reported in Figs 4 and 5. Trp at the C-terminus is more solvent accessible and exists in a relatively less apolar environment as compared to the N-terminus homolog. According to these results, the Trp at the N-terminus is likely to bind to the polar–apolar interface via electrostatic binding between the positively charged Lys and the negatively charged lipids. These measurements hint that the C-terminus may reside in the aqueous phase. Insertion of hydrophobic residues into the lipid bilayer is generally accompanied by a decrease in mem- brane fluidity and a corresponding increase in the anisotropy constant. Such changes are typically observed after insertion of hydrophobic peptides into the membrane [43]. For example, functional mutant OmpA signal peptides that possess high hydrophobic contents insert into membranes and increase DPH anisotropy [44]. Similarly, a peptide fragment from the cytotoxic protein a-sarcin penetrates into the hydro- phobic core of the bilayer and substantially increases DPH anisotropy at a temperature above the phase transition [45]. In contrast, the opposite was observed here, i.e. decreased anisotropy constant and increased mem- brane fluidity, possibly because of membrane destabili- zation by the formation of b-aggregates. This observation is consistent with a recent report that the A B C D E F 300 300 250 200 150 100 250 600 500 400 300 200 100 0 200 150 180 350 300 250 200 150 100 50 0 160 140 120 100 80 60 40 20 0 100 50 0 50 0 350 400 450 300 600 500 400 300 200 100 0 400 500 300 350 400 450 300 400 500 300 350 400 450 300 400 500 Wavelength (nm) WT/W@36 + PG Dan–WT/W@36 + PG Dan–A30G/W@36 + PG Dan–A30l/W@36 + PG A30G/W@36 + PG A30l/W@36 + PG Wavelength (nm) Wavelength (nm) Wavelength (nm) Wavelen g th (nm) Wavelen g th (nm) Fluorescence intensity (a.u.) Fluorescence intensity (a.u.) Fluorescence intensity (a.u.) Fluorescence intensity (a.u.) Fluorescence intensity (a.u.) Fluorescence intensity (a.u.) 0 h 24 h 48 h 72 h 96 h 120 h 0 h 24 h 48 h 72 h 96 h 120 h 0 h 24 h 48 h 72 h 96 h 120 h 0 h 24 h 48 h 72 h 96 h 120 h 0 h 24 h 48 h 72 h 96 h 120 h 0 h 24 h 48 h 72 h 96 h 120 h Fig. 8. Effect of PG lipids on the time dependence of FRET for the dansyl-Trp peptide derivatives dansyl-WT ⁄ W36, dansyl- A30G ⁄ W36, and dansyl-A30I ⁄ W36 (D–F), and for their negative controls WT ⁄ W36, A30G ⁄ W36, and A30I ⁄ W36 (A–C). Spectra were acquired at different time intervals (0, 24, 48, 72, 96 and 120 h) by exciting Trp at 280 nm and recording the emission between 300 and 535 nm at 298 K. Interactions of Ab(28–35) and its mutants with lipids S. Nagarajan et al. 2422 FEBS Journal 275 (2008) 2415–2427 ª 2008 The Authors Journal compilation ª 2008 FEBS membrane ultrastructure is similarly disrupted by islet amyloid polypeptide and polymyxin B [46], Ab in the presence of Golgi bilayers [47], transthyretin amyloid binding to the plasma membrane, and aggregated Ab on the synaptic plasma membrane [48]. Our results are in agreement with the study of Terzi et al. (1997) and are in partial agreement with Dante et al. [49]. In that case, Ab(28–35) bound electrostatically to the nega- tively charged membrane only under physiological con- ditions. From the above results, we propose that the peptide N-terminus interacts with the negatively charged lipids, whereas the C-terminal portion is oriented away from the membrane surface. This association must result in close intermolecular contact between the hydrophobic residues of the peptide. Hence, the kinetic barriers for the association of peptides into aggregates are greatly reduced by the binding of the peptides to the mem- brane surface. This study is supported by previous studies showing an immediate increase in membrane disruption when soluble Ab(25–35) was added to nega- tively charged membranes [50]. As membrane fluidity is known to be important for normal cell function and viability [51], this phenomenon of membrane disrup- tion ⁄ destabilization may be a crucial mechanism of amyloid neurotoxicity. Some studies suggest that the b-amyloid peptides bind electrostatically only to the polar head groups, i.e. do not become embedded within the hydrophobic interior [52]. Our results pro- vide firm evidence that these smaller peptides bind to negatively charged lipids through electrostatic interac- tions and disturb the membrane by forming a b-sheet scaffold at the membrane–water interface. In summary, our results indicate that electrostatic interactions are responsible for the initial binding of negatively charged lipids and positively charged pep- tides. We conclude from this comparison that the b-sheet preferences that were observed for the peptides in negatively charged lipid depends on the intrinsic b-sheet propensities, and side-chain–side-chain and side-chain–backbone interactions. This study also gives an understanding of the specific role played by hydro- phobic residues in membrane lipid binding and can be exploited for the development of specific therapeutic drugs to prevent amyloid peptide neuronal membrane toxicity. Experimental procedures Peptide synthesis and characterization All Fmoc amino acids were purchased from Nova Biochem (San Diego, CA, USA). Pentafluorophenol was obtained from Spectrochem Ltd (Bombay, India). Wang resin was purchased from SRL Ltd (Bombay, India). All analytical grade organic solvents used in the present study were pro- cured from Merck Ltd (Bombay, India) and S.D. Fine chemicals (Bombay, India). Ab(28–35) and the A30G and A30I peptides (Table 1) were synthesized manually by stan- dard solid-phase synthesis using Wang resin and amino acids protected by the pentafluorophenyl ester of Fmoc, as previously reported [53]. For the peptides labeled with Trp, Trp was added at positions 27 and 36, respectively, using the same procedure. To synthesize compounds, addition of a dansyl group to the N-terminus of the WT ⁄ W36, A30G ⁄ W36 and A30I ⁄ W36 peptides was effected by incu- bating the protected, resin-bound peptide with 1.5 equiva- lents of dansyl chloride and three equivalents of triethylamine for 45 min at room temperature. The resin was washed several times with dichloromethane and dried under vacuum. A Kaiser test was performed to check the completion of the reaction. The peptides were purified by HPLC and characterized by 500 MHz proton NMR A B C 600 500 400 Tryptophan maximum fluorescence Intensity (a.u.) Tryptophan maximum fluorescence Intensity (a.u.) Tryptophan maximum fluorescence Intensity (a.u.) 300 200 0 50 Time (h) Time (h) Time (h) 100 WT/W@36 WT+PG/W@36 A30G/W@36 A30l+PG/W@36 A30l/W@36 A30G+PG/W@36 150 0 50 100 150 0 50 100 150 100 0 600 450 400 350 300 250 200 150 100 50 0 500 400 300 200 100 0 Fig. 9. Time dependence of the Trp maximum emission fluorescence for the WT ⁄ W36 (A), A30G ⁄ W36 (B) and A30I ⁄ W36 (C) peptides in the absence (r) and presence ( ) of PG vesicles. Spectra were acquired at different time intervals (0, 24, 48, 72, 96 and 120 h). Trp-labeled peptides were excited at 280 nm, and the emission was measured from 300 to 500 nm. S. Nagarajan et al. Interactions of Ab(28–35) and its mutants with lipids FEBS Journal 275 (2008) 2415–2427 ª 2008 The Authors Journal compilation ª 2008 FEBS 2423 spectroscopy. The purity of the peptides obtained was 88%. Furthermore, the molecular masses of the peptides were confirmed by MALDI-TOF MS and compared with theo- retical molecular masses (Table 1). Liposome preparation Lyophilized PG and PC lipids (Sigma-Aldrich) were dis- solved in 3 : 1 chloroform ⁄ methanol, and then left to dry in air and then in vacuum for 6 h to remove any traces of solvent. Millipore water was added to the lipid film and sonicated using an ultrasonicator bath, until an optically clear solution was obtained. The phosphate concentration was determined by the method of Ames [54]. Peptide dissolution WT, A30G and A30I peptides were pretreated with trifluo- roacetic acid ⁄ trifluoroethanol as previously described [55]. Briefly, peptides were incubated with trifluoroacetic acid ⁄ trifluoroethanol (1 : 5) for 2 h. Then, trifluoroacetic acid ⁄ trifluoroethanol was removed under a stream of nitro- gen gas. This was followed by the immediate addition of 10% acetic acid in water, sonication, and lyophilization. The lyophilized peptides were dissolved in NaCl ⁄ P i and used immediately for the studies carried out in this work. CD Lipid ⁄ peptide ratios were maintained at 30 : 1, with final peptide concentrations of 50 lm. The effect of various lip- ids on peptide conformation was determined by adding an aliquot of freshly prepared peptide stock solution to pre- formed lipid vesicles under continual stirring. The contribu- tion of lipid vesicles to the CD signal was removed by subtracting the CD spectra of pure lipid vesicles from the CD spectra of peptide ⁄ lipid suspensions. Spectra were acquired by means of a JASCO J-715 spectropolarimeter (Jasco, Tokyo, Japan) equipped with a thermostated cell holder, using a quartz cell of 1 mm path length at 25 °C. This equipment was calibrated using ammonium-d 10 -cam- phor sulfonic acid as recommended by the instrument man- ufacturer. WT, A30G and A30I peptides were dissolved in 10 mm NaCl ⁄ P i (pH 7.4), at a concentration of 50 lm. The CD quartz cell was placed near the photomultiplier tube to reduce the scattering from the lipid vesicles. Spectra were collected over the wavelength range 260–190 nm and smoothed from the buffer spectra. The CD value was expressed as molar ellipticity. AFM The samples were imaged with a Shimadzu-5500 atomic force microscope (Shimadzu, Kyoto, Japan), using tapping mode scanning and an Si 3 N 4 tip. The tube scanner was a 30 lm scan master. The images shown were taken in the noncontact AFM imaging mode. Samples were prepared for AFM imaging by drying a 10 lL sample from the reac- tion mixture on freshly cleaved mica with nitrogen gas. The buffer was washed from the surface of the mica with dou- ble-distilled water, and the mica was dried again. Steady-state fluorescence anisotropy Anisotropy experiments were performed on a Perkin Elmer fluorimeter equipped with manual polarizers. Excitation and emission wavelengths were set at 360 and 425 nm, with slit widths of 1 and 4 nm, respectively. Our system was ini- tially calibrated using DPH in mineral oil, which should give an anisotropy equal to 1. The g-factor was calibrated using horizontally polarized excitation and subsequent com- parison of the horizontal and vertical emissions, which for our machine is 0.88. Lipid vesicles were diluted to 500 lm with NaCl ⁄ P i , incubated for 20–30 min in the presence and absence of Ab, and then incubated for a further 30 min with DPH at a 1 : 500 probe ⁄ lipid ratio. Fluorescence intensity was measured with the excitation polarizer in (I vv ) and horizontal (I vh ) positions, and anisotropy, r, was calcu- lated using Eqn (1) [56]: r ¼ I vv À gI vh =I vv þ 2gI vh ð1Þ Lipid vesicles in the absence of DPH were measured in order to evaluate the effect of light scattering on our mea- surements. Fluorescence measurements Intrinsic fluorescence The kinetics of aggregation were monitored for peptides N-labeled and C-labeled with Trp by exciting at 280 nm and detecting between 300 and 540 nm both in the absence and in the presence of PG and PC vesicles. Peptides were used at 10 lm in 10 mm NaCl ⁄ P i (pH 7.4). Measurements were performed on a Perkin Elmer LS 45 fluorimeter equipped with a xenon lamp and a thermostatically con- trolled cuvette holder using a semi-microquartz cuvette (1 cm path length; excitation and emission bandpass of 2 nm). Spectra were plotted, and the wavelength and inten- sity at the maximum emission were recorded. All the fluo- rescence studies were carried out at 25 °C. Acrylamide quenching Acrylamide was added to the Trp-labeled peptide solutions, both in the absence and in the presence of PG and PC vesi- cles. Fluorescence intensities were corrected for dilution effects. Fluorescence quenching data were analyzed using the general form of the Stern–Volmer equation (Eqn 2) Interactions of Ab(28–35) and its mutants with lipids S. Nagarajan et al. 2424 FEBS Journal 275 (2008) 2415–2427 ª 2008 The Authors Journal compilation ª 2008 FEBS [...]... Interactions of Ab(28–35) and its mutants with lipids I0 =I ¼ ð1 þ Ksv ½QŠÞ ð2Þ where Io and I are the fluorescence intensities in the absence and presence of the quencher, respectively, Ksv is the dynamic quenching constant, and [Q] is the quencher concentration The Ksv value was determined by titrating acrylamide from a 5 mm stock solution to 10 lm peptide in either the presence or the absence of lipids... Dansyl-WT ⁄ W36 and WT ⁄ W36, dansyl -A30G ⁄ W36 and A30G ⁄ W@36, and dansyl -A30I ⁄ W36 and A30I ⁄ W36 peptide pairs were used to study folding and unfolding of lipidfree and lipid-bound peptides Intramolecular FRET was minimized by mixing the labeled peptides with nonlabeled peptides at a molar ratio of 1 : 8 (total 9 lm) in 10 mm NaCl ⁄ Pi Emission spectra were recorded between 300 and 550 nm, using... neurotoxic fragment of the b -amyloid peptide Biochem Biophys Res Commun 202, 1142–1148 Pillot T, Goethals M, Vanloo B, Talussot C, Brasseur R, Vandekerckhove J, Rosseneu M & Lin L (1996) Fusogenic properties of the C-terminal domain of the Alzheimer b -amyloid peptide J Biol Chem 271, 28757– 28765 Jobling MF, Stewart LR & White AR (1999) The hydrophobic core sequence modulates the neurotoxic and secondary... (2006) The stability of monomeric intermediates controls amyloid formation: Abeta25–35 and its N27Q mutant Biophys J 90, 3365–3374 Del Mar Martinez-Senac M, Villalain J & GomezFernandez JC (1999) Structure of the Alzheimer beta -amyloid peptide (25–35) and its interaction with negatively charged phospholipid vesicles Eur J Biochem 265, 744–753 Terzi E, Holzemann G & Seelig J (1995) Self-association of b -amyloid. .. solution and binding to lipid membranes J Mol Biol 252, 633–642 Jarrett JT, Berger EP & Lansbury PT Jr (1993) The carboxy terminus of the beta amyloid protein is critical for the seeding of amyloid formation: implications for the pathogenesis of Alzheimer’s disease Biochemistry 32, 4693–4697 Pillot T, Drouet B, Queille S, Labeur C, Vandekerchkhove J, Rosseneu M, Pincon-Raymond M & Chambaz J (1999) The nonfibrillar... The isolation and characterization of a novel corticostatin ⁄ defensin-like peptide from the kidney J Biol Chem 271, 654–659 21 Pike CJ, Walencewicz-Wasserman AJ, Kosmoski J, Cribbs DH, Glabe CG & Cotman CW (1995) Structure–activity analyses of beta -amyloid peptides: contri- FEBS Journal 275 (2008) 2415–2427 ª 2008 The Authors Journal compilation ª 2008 FEBS 2425 Interactions of Ab(28–35) and its mutants. .. (2006) Effect of ionic strength on the organization and dynamics of membrane-bound melittin Biophys Chem 124, 115–124 Hasselbacher CA, Rusinova E, Waxman E, Rusinova R, Kohanski RA, Lam W, Du Guha AJ, Lin TC & Polikarpov I (1995) Environments of the four tryptophans in the extracellular domain of human tissue factor: comparison of results from absorption and fluorescence difference spectra of tryptophan... Pepys MB (1994) Amyloidosis Histopathology 25, 403–414 5 Yankner BA, Dawes LR, Fisher S, Villa-Komaroff L, Oster-Granite ML & Neve RL (1989) Neurotoxicity of a fragment of the amyloid precursor associated with Alzheimer’s disease Science 245, 417– 420 6 Pike CJ, Burdick D, Walencewicz AJ, Glabe CG & Cotman CW (1993) Neurodegeneration induced by beta -amyloid peptides in vitro: the role of peptide assembly... for helping with the DPH anisotropy study References 1 Selkoe DJ (2006) Amyloid beta-peptide is produced by cultured cells during normal metabolism: a reprise J Alzheimers Dis 9, 163–168 2 Hardy J & Selkoe DJ (2002) The amyloid hypothesis of Alzheimer’s disease: progress and problems on the road to therapeutics Science 297, 353–356 3 Selkoe DJ (1999) Translating cell biology into therapeutic advances... intercalated in anionic and zwitterionic lipid membranes to different extents Biophys J 83, 2610–2616 Mattson MP, Begley JG, Mark RJ & Furukawa K (1997) Abeta25–35 induces rapid lysis of red blood cells: contrast with Abeta1–42 and examination of underlying mechanisms Brain Res 771, 147–153 Interactions of Ab(28–35) and its mutants with lipids 51 Simons K & Toomre D (2000) Lipid rafts and signal transduction . study, we designed and tested the properties of the amyloid core fragment Ab(28–35) and two of its mutants, A30G and A30I, to understand the peptide–membrane. Lipid-induced conformational transition of the amyloid core fragment Ab(28–35) and its A30G and A30I mutants Sureshbabu Nagarajan 1 ,