Skip to main content
NCBI home page
Protein Science : A Publication of the Protein Society logoLink to Protein Science : A Publication of the Protein Society
. 2004 May;13(5):1251–1259. doi: 10.1110/ps.03442404

Stabilization of discordant helices in amyloid fibril-forming proteins

Anna Päiviö 1, Erik Nordling 2, Yvonne Kallberg 2, Johan Thyberg 3, Jan Johansson 1
PMCID: PMC2286751  PMID: 15096631

Abstract

Several proteins and peptides that can convert from α-helical to β-sheet conformation and form amyloid fibrils, including the amyloid β-peptide (Aβ) and the prion protein, contain a discordant α-helix that is composed of residues that strongly favor β-strand formation. In their native states, 37 of 38 discordant helices are now found to interact with other protein segments or with lipid membranes, but Aβ apparently lacks such interactions. The helical propensity of the Aβ discordant region (K16LVFFAED23) is increased by introducing V18A/F19A/F20A replacements, and this is associated with reduced fibril formation. Addition of the tripeptide KAD or phospho-L-serine likewise increases the α-helical content of Aβ(12–28) and reduces aggregation and fibril formation of Aβ(1–40), Aβ(12–28), Aβ(12–24), and Aβ(14–23). In contrast, tripeptides with all-neutral, all-acidic or all-basic side chains, as well as phosphoethanolamine, phosphocholine, and phosphoglycerol have no significant effects on Aβ secondary structure or fibril formation. These data suggest that in free Aβ, the discordant α-helix lacks stabilizing interactions (likely as a consequence of proteolytic removal from a membrane-associated precursor protein) and that stabilization of this helix can reduce fibril formation.

Keywords: α-helix, protein structure, protein aggregation, conformational disease

Amyloid fibrils formed from specific proteins are associated with ~20 different diseases, including Alzheimer’s disease (AD; amyloid β-peptide [Aβ] forms fibrils), prion diseases (prion protein [PrP] forms fibrils), hereditary systemic amyloidosis (lysozyme mutants form fibrils), and familial amyloid polyneuropathy (transthyretin forms fibrils; Kelly 1998). Amyloid fibrils are composed of polypeptide chains in β-strand conformation, which form β-sheets running perpendicular to the long axis of the fibril (Serpell 2000). The morphology and molecular architecture of amyloid fibrils are apparently very similar, although they are formed from proteins with widely different native structures, sizes, and localization (Dobson 1999). Moreover, under partly denaturing conditions, fibrils can be formed from almost any protein, suggesting that the polypeptide backbone is a main determinant of the fibril structure (Fändrich et al. 2001; Srisailam et al. 2003). Although many polypeptide sequences thus can form fibrillar β-sheet structure under certain conditions, the ability to do so under physiological conditions is apparently limited to a small number of proteins. Out of ~1300 nonhomologous proteins with experimentally determined three-dimensional structures, PrP, Aβ, and ~30 other proteins were recently found to contain α-helices composed of amino acid sequences that are strongly predicted to form β-strands, called discordant helices (Kallberg et al. 2001). This observation raises the possibility that conflicts in structural preferences localized to comparatively short linear polypeptide regions can underlie fibril formation. In agreement with these results, it was recently found that the discordant helix of PrP is frustrated in its helical state (Dima and Thirumalai 2002). Four of the 10 proteins with the longest discordant helices (≥11 residues) have been analyzed, and all form fibrils. It has also been observed that by mutating the amino acid sequences of the discordant helices of Aβ and lung surfactant protein C, resulting in predicted helical structures instead, fibril formation can be abrogated (Kallberg et al. 2001; Hosia et al. 2002). Because discordant helices are made from amino acid sequences that appear intrinsically unsuited for helix formation, it is likely that these helices require additional interactions for stabilization. In this study, we analyzed the 38 longest discordant helices found previously (Kallberg et al. 2001) regarding their native local environments, and found that all of them except Aβ interact with polypeptide segments or lipid membranes.

The 39–43 residues-long (preferentially 40 or 42 residues) Aβ is invariably present in amyloid plaques found in association with AD, and formation of Aβ fibrils is thought to be part of the cause of this devastating disease (Selkoe 2000). The postmortem amounts of Aβ in the cerebral cortices of AD patients correlate with the disease progression, indicating that Aβ is a rational therapeutic target (Näslund et al. 2000). Aβ is generated by proteolytic cleavages of a large transmembrane protein, the amyloid precursor protein (APP). The transmembrane helix of APP is predicted to start at a position corresponding to residue 29 in Aβ. Cleavage by β-secretase generates the N terminus of the Aβ part and leaves behind a C-terminal membrane-associated stub, which is cleaved by γ-secretase, generating free Aβ. Cleavage by α-secretase C-terminally of residue 16 in Aβ generates a nonamyloidogenic peptide (Selkoe 1999; Esler and Wolfe 2001). Hereditary forms of early onset AD are associated with point mutations in three regions of APP. Two of these regions are located in close vicinity of the N- and C-terminal ends of Aβ, and are likely associated with elevated production of Aβ(1–40) and/or the more amyloidogenic variant Aβ(1–42). The third region covers positions Ala 21, Glu 22, and Asp 23 of Aβ, and the pathogenic mechanisms associated with mutations of these residues are not entirely understood (Haass and Steiner 2001).

The secondary structure of Aβ(1–40/42) peptides in aqueous solution is mainly disordered (Riek et al. 2001), whereas in the presence of 20%–40% (v/v) trifluoroethanol (TFE), a significant portion of helical structure (~15%–30%) is found (Barrow and Zagorski 1991; Soto et al. 1995; Sticht et al. 1995). NMR structure determinations of Aβ(1–40) and Aβ(1–42) in SDS micelles revealed an α-helix covering positions 15–36 with a kink located at positions 25–27 (Coles et al. 1998), or two helices covering positions 10–24 and 28–42, respectively (Shao et al. 1999). In 40% (v/v) TFE two helices, covering positions 15–23 and 31–35, were found (Sticht et al. 1995). Aβ(1–42) is more prone to aggregate and form plaques than Aβ(1–40). This is not reflected in different conformations of the monomeric peptides (Riek et al. 2001), but may be attributed to the presence of two additional unpolar residues (Ile–Ala) at the C-terminal end, making the Aβ(1–42) less soluble in aqueous solvents (Jarrett et al. 1993). The unpolar C-terminal region of Aβ, which emanates from the predicted transmembrane part of APP, thus possibly influences plaque formation by decreasing Aβ solubility and increasing peptide–peptide contacts. However, Aβ(1–28) also forms fibrils (Kirschner et al. 1987; Tjernberg et al. 1996), indicating that the hydrophobic C-terminal region of Aβ is not essential for fibril formation. Several lines of evidence indicate that a region centering around positions 17–20 is important for Aβ fibril formation. In Aβ(1–40), positions 16–20 were found to be involved in formation of Aβ intermolecular contacts and fibril formation, and removal of Aβ positions 14–23 prevents fibril formation (Tjernberg et al. 1996, 1999). In Aβ(1–42), destabilization of a helix covering residues 11–24, in particular residues 17–24, is critical for α-helix →β-strand conversion and fibril formation (Janek et al. 2001). The discordant helix of Aβ covers residues K16LVFFAED23. Herein we found that residue replacements in this region and compounds that stabilize Aβ-helical conformation reduce fibril formation.

Results

Native environments of discordant helices

The interactions with surrounding amino acid residues were determined for each residue of the 38 longest discordant helices within their native proteins. For this purpose, we used the relative accessibility as a measure of the degree of shielding imposed by surrounding residues. The result shows that the helices fall into one of three categories (Fig. 1 ▶). One group represents helices that are predominantly buried with exception of the terminal regions (Fig. 1A ▶). A second group represents helices with periodic variations in their residue accessibilities. The periodicity corresponds to three to four residues, and ~50% of the residues in these helices are buried (Fig. 1B ▶), corresponding to helices located at the protein surface. For both these groups, it is concluded that the local protein environment stabilizes the helices. A third group represents helices that are apparently exposed along their entire length (Fig. 1C ▶). These six helices would be expected to adopt a β-strand conformation as predicted from their sequences. However, in three cases (1bct, bacteriorhodopsin; 1spf, surfactant-associated protein C; and 1bl1, parathyroid hormone receptor), the helices are buried in a lipid membrane in their native environment, and in two cases (1aa0, fibritin deletion mutant; and 2ifo, inovirus major coat protein), the discordant helices are involved in oligomer formation. Interactions with surrounding lipids or other polypeptides caused by oligomerization are expected to stabilize these helices. The only remaining case of the 38 studied discordant helices for which no local interactions are apparent is thus 1ba6 (human Aβ). Possibly, lack of stabilization of helical Aβ contributes to its tendency to form β-sheet aggregates and fibrils. We therefore investigated whether stabilization of the discordant Aβ-helix, by residue replacements or by addition of ligands, affects fibril formation.

Figure 1.

Figure 1.

Open in a new tab

Surface accessibility of discordant helices. The relative solvent-accessible surface of each discordant helix residue in the respective protein calculated as described in Materials and Methods. For each protein, the PDB code and the trivial name are given (see Kallberg et al. 2001 for details). (A) Helices with mainly buried residues. (1aur) Carboxylesterase, (1b2v) heme-binding protein A, (1b5e) dCMP hydroxymethylase, (1b8o) purine nucleoside phosphorylase, (1cpo) chloroperoxidase, (1ggt) coagulation factor XIII, (1h2a) hydrogenase, (1iab) astacin, (1jkm) brefeldin A esterase, (1lml) leishmanolysin, (1mhd) smad MH1 domain, (1mnm) transcription factor MVM1 (middle of helix is not discordant and not plotted), (1mty) methane monooxygenase (middle of helix is not discordant and not plotted), (1nom) DNA polymerase β, (1noz) DNA polymerase, (1qut) lytic transglycosylase Slt35, (1sra) osteonectin, (1tah) lipase, (1tca) lipase B, (1vns) chloroperoxidase (middle of helix is not discordant and not plotted), (1wer) Ras-GTPase-activating domain of p120GAP, (2sqc) squalene-hopene cyclase, (3aig) adamalysin II, (3pte) transpeptidase. (B) Helices with at least two exposed residues (accessibility ratio >0.3) separated by buried residues. The first buried residue in each helix is aligned. (1b10) Prion protein, (1cv8) staphopain, (1ecr) replication terminator protein, (1kpt) killer toxin, (1pbv) sec7 domain of exchange factor ARNO, (1wer) Ras-GTPase-activating domain of p120GAP, (2erl) pheromone Er-1, (2occ) cytochrome C oxidase. (C) Helices with mainly exposed residues. (1aa0) Fibritin deletion mutant, (1ba6) amyloid β-peptide (Aβ), (1bct) bacteriorhodopsin, (1bl1) parathyroid hormone receptor, (1spf) surfactant-associated protein C, (2ifo) inovirus major coat protein.

Aβ discordant helix stabilization by residue replacements

The central three-residue segment Val–Phe–Phe in the discordant Aβ-helix was replaced with Ala–Ala–Ala. To minimize possible influences on secondary structure and fibril formation from other parts of Aβ, a peptide corresponding to Aβ(12–28) was studied. The residue replacements yielded a peptide that is predicted to form α-helical structure; that is, the discordant nature is abolished (Fig. 2A ▶). The secondary structures of Aβ(12–28) and Aβ(12–28; V18A/F19A/F20A) were analyzed by CD spectrometry. This showed that the Ala-substituted peptide has a higher propensity to form helical structure, as estimated from the residual molar ellipticity at 222 nm in the presence of 30% and 70% TFE (Fig. 2B ▶). Aβ(12–28) forms amyloid fibrils, but Aβ(12–28; V18A/F19A/F20A) does not form any detectable fibrils, as judged by the ThT fluorescence of the two peptides (Fig. 2C ▶). Analysis by electron microscopy confirmed that Aβ(12–28) forms fibrillar structures (see also Tjernberg et al. 1996), but for Aβ(12–28; V18A/F19A/ F20A) no fibrils were observed. These results show that increasing the helical propensity of Aβ(12–28) by residue replacements abrogates fibril formation. Differences in fibril formation between Aβ(12–28) and Aβ(12–28; V18A/ F19A/F20A) are observed in phosphate buffer, although their secondary structures in buffer are very similar, as judged by CD spectroscopy; differences only become evident in the presence of TFE. This suggests that small shifts in secondary structure populations in solution may have significant effects on fibril formation. However, as suggested by molecular dynamics simulations of Aβ(16–22; L17S/F19S/F20S; Klimov and Thirumalai 2003), altered peptide–peptide contacts as an effect of the side-chain replacements between Aβ(12–28) and Aβ(12–28; V18A/ F19A/F20A) cannot be ruled out. Therefore, ligands that can stabilize Aβ-helical conformation were searched for.

Figure 2.

Figure 2.

Open in a new tab

Removal of the Aβ discordant nature prevents fibril formation. (A) Sequences and predicted secondary structures for Aβ(12–28) and Aβ(12–28; V18A/F19A/F20A). Blue cylinders (bottom) represent the experimentally determined central helix of Aβ; the central yellow strand or blue cylinder represents β-strand or helical structure, respectively, predicted by the PHD method in which the figures give the reliability indexes (range 1–9); and the top line E or H represents extended or helical structure, respectively, predicted using the Chou-Fasman algorithm (see Kallberg et al. 2001 for details). In contrast to the native sequence, a helical structure is predicted for the sequence containing the V18A/F19A/F20A replacements. (B) Residual molar ellipticity at 222 nm as a function of TFE content for Aβ(12–28) (open squares) and Aβ(12–28; V18A/F19A/F20A) (filled triangles). (C) ThT fluorescence of solutions containing 100 βM Aβ(12–28) (open squares) or Aβ(12–28; V18A/F19A/F20A) (filled triangles) during 14 d of incubation in 10 mM sodium phosphate buffer (pH 6.0) at 37°C.

Aβ discordant helix stabilization by ligands

Addition of 1 mM of the tripeptide KAD to a solution of 100 μM Aβ(12–28) in 40% TFE results in increased helical structure as determined by CD spectroscopy (Fig. 3 ▶). From the residual molar ellipticity at 222 nm, a helical content of 13% is estimated for Aβ(12–28), and addition of KAD results in 18% helix. In contrast, addition of the tripeptides AAA (Fig. 3 ▶), FRF, KKK, or DDD gives no significant effect on the Aβ(12–28) CD spectrum. Addition of KAD to Aβ(12–28) in TFE-free buffer gives no detectable change in the CD spectrum.

Figure 3.

Figure 3.

Open in a new tab

KAD increases the helical content of Aβ(12–28). CD spectra of 100 μM Aβ(12–28) in 40% aqueous TFE (open squares) and after addition of 1 mM KAD tripeptide (upper graph) or 1 mM AAA (lower graph), after subtraction of the contribution from respective tripeptide. The residual molar ellipticity is expressed in kilodegrees centimeters squared per decimole.

Analysis of Aβ fibril formation in the presence of oligopeptides by ThT fluorescence did not yield consistent results, possibly because of interference due to the relatively high concentrations of oligopeptides. The amounts of fibrils formed, after 3 d of incubation at 37°C, from 100 μM of Aβ(14–23), Aβ(12–24), and Aβ(1–40) peptides in the absence or presence of 1 mM of different tri- or tetrapeptide ligands, were therefore determined by electron microscopy. In these experiments, the density of fibrils with a morphology similar to that of fibrils formed from the Aβ peptides alone was determined. Figure 4 ▶ exemplifies the extent of Aβ(1–40) fibrils formed in the presence of AAA and KAD, respectively. For all three Aβ peptide variants studied, a substantial reduction of fibril density was observed in the presence of the KAD tripeptide, but not in the presence of FRF, AAA, KKK, or DDD tripeptides (Fig. 5 ▶). Likewise, KAD, but not AAA, reduced fibril formation of Aβ(12–28) (data not shown). Acetyl-KAD-amide was found to be equally efficient in reducing Aβ(1–40) fibril formation as the peptide with free termini, and both AAA and acetyl-AAA-amide showed only marginal effects on Aβ(1–40) fibrillation (Fig. 5 ▶). Replacing the central Ala with Phe resulted in a tripeptide (KFD) that reduced Aβ(14–23) fibril formation, but to a lesser extent than KAD (Fig. 6 ▶). Extending the length of the dipolar peptides with one residue resulted in no reduction of Aβ(14–23) fibrillation; the tetra-peptide KFFE even seemed to promote fibril formation slightly (Fig. 6 ▶). Finally, the effects of KAD, AAA, and KFFE on the time-dependent aggregation of Aβ(1–40) were determined from the amount of Aβ peptide left in solution after 20,000g centrifugation. In the absence of oligopeptides or in the presence of AAA, Aβ(1–40) aggregates completely in ~15 d; in the presence of KAD, aggregation takes 40 d; and in the presence of KFFE it takes ~5 d (Fig. 7 ▶). These results agree with the results obtained from the electron microscopy studies regarding the relative effects of KAD, AAA, and KFFE on Aβ fibrillation. The apparent increase in Aβ fibril formation and aggregation after addition of KFFE may be explained by the recent finding that this peptide forms β-strand structure and aggregates into amyloid fibrils on its own (Tjernberg et al. 2002).

Figure 4.

Figure 4.

Open in a new tab

KAD and AAA effects on Aβ fibril formation. Electron micrographs of fibrillar material formed from 100 μM Aβ(1–40) in the presence of 1 mM AAA (A) or KAD (B).

Figure 5.

Figure 5.

Open in a new tab

Reduction of Aβ fibril formation by tripeptides. Here 100 μM Aβ(14–23), Aβ(12–24), or Aβ(1–40) in sodium phosphate-buffered saline was incubated for 3 d at 37°C either alone or in the presence of 1 mM each indicated tripeptide. After incubation, the solutions were centrifuged at 20,000g and the number of fibril bundles in the pellet was estimated by electron microscopy. Unless indicated otherwise, the tripeptides used have free N and C termini. The mean values and standard deviations of three independent experiments are shown.

Figure 6.

Figure 6.

Open in a new tab

Reduction of Aβ(14–23) fibril formation by tri- and tetrapeptides. Here 100 μM Aβ(14–23) was incubated in the presence of 1 mM the indicated tri- or tetrapeptides, and amounts of fibrils formed were determined as described for Figure 5 ▶.

Figure 7.

Figure 7.

Open in a new tab

Effects on Aβ(1–40) aggregation of KAD, AAA, and KFFE. Here 100 μM Aβ(1–40) was incubated in phosphate-buffered saline in the presence of 1 mM KAD, AAA, or KFFE at 37°C. At the indicated time points, aliquots of the solutions were withdrawn and centrifuged at 20,000g for 20 min, and the relative amounts of Aβ(1–40) peptide in the supernatants were determined by amino acid analysis. The values shown are the mean of duplicate samples.

The effects of phospho-L-serine, phosphocholine, phosphoethanolamine, and phosphoglycerol on Aβ-helical content and fibril formation are summarized in Figure 8 ▶. Phos-pho-L-serine, but not the other phosphate-containing compounds analyzed, increases the helical content of Aβ(12–28) in 30% TFE and reduces fibril formation. Estimation from the ellipticity at 222 nm yields 17% helical content for Aβ(12–28) and ~22% in the presence of 10-fold molar excess of phospho-L-serine. None of the compounds gave any effect on the CD spectrum of Aβ(12–28) in TFE-free buffer. The results obtained with KAD and phospho-L-serine indicate that small increases in Aβ-helical propensity, measurable by CD spectroscopy only in the presence of TFE, results in reduced fibril formation.

Figure 8.

Figure 8.

Open in a new tab

Phospho-L-serine increases helical content and reduces fibril formation. CD spectra of 100 μM acetyl-Aβ(12–28)-amide in 30% aqueous TFE alone (open squares), and after addition of 1 mM phospho-L-serine (filled triangles, A) or 1 mM phosphoethanolamine (filled circles, B). Spectra in the presence of phosphocholine or phosphoglycerol were virtually identical to that of Aβ(12–28). (C) ThT fluorescence of solutions containing 100 μM Aβ(12–28) alone (open squares), Aβ(12–28) plus 1 mM phospho-L-serine (filled triangles), or Aβ(12–28) plus 1 mM phosphoethanolamine (filled circles) during 14 d of incubation at 37°C.

Discussion

Discordant helices are predicted to form β-strands, and are found in proteins that can undergo α-helix → β-strand conversion and form amyloid fibrils (Kallberg et al. 2001; Thirumalai et al. 2003). Herein we have looked for factors that can stabilize such helices. In most of the studied cases, the discordant helix interacts closely with other regions of the native protein or a lipid membrane (Fig. 1 ▶), which is expected to stabilize their helical conformations and thereby reduce the tendency to form β-strands. These findings indicate that inherent secondary structure propensities cannot solely account for the conformation of a given stretch of amino acids, but that other factors such as the local environment must also be taken into consideration. Discordant helices show conflicts in secondary structures dictated by local amino acid sequence preferences and those dictated by the tertiary structure or by protein–membrane interactions. As a consequence, it is likely that discordant helices are not formed until their local protein environment has attained a near-native structure (or until they are inserted into a lipid membrane). A recent mutational study of acylphosphatase showed that regions responsible for initiating the process of aggregation do not participate in the establishment of the folding nucleus (Chiti et al. 2002). The proposed late folding of discordant helices is in line with the finding that of the seven transmembrane regions in bacteriorhodopsin, only the most C-terminal helix (which is discordant) forms a hyperstable β-sheet aggregate in isolation, whereas five of the other regions form stable transmembrane helices in isolation (Hunt et al. 1997). The authors pointed out that C-terminally located helices may have a lower intrinsic stability, and their efficient folding requires rapid folding of the segments that are synthesised first, so that they can serve as folding templates for the distal segments (Hunt et al. 1997). Perhaps the requirement of a folding template applies to discordant helices in general.

The Aβ discordant helix, covering positions 16–23, differs from the other helices now studied in that it lacks apparent stabilizing interactions. This likely reflects that Aβ is a proteolytic fragment of APP. Aβ positions 16–23 show helical structure in the presence of membrane-mimicking solvents or detergents, suggesting that this region is helical also in membrane-associated APP. However, for liberated Aβ in aqueous solution, mainly unordered conformation is detected by spectroscopic methods (Serpell 2000; Riek et al. 2001). In the absence of helix-stabilizing interactions, the inherent propensity of the Aβ discordant region to form β-sheet structure is expected to contribute to the ability to form fibrils. Mutating Lys 16, Leu 17, and Phe 20 to Ala changes the secondary structure propensity of Aβ(1–28) so that the discordant nature is abolished (Kallberg et al. 2001) and prevents fibril formation (Tjernberg et al. 1996). Like-wise, the replacement Val18Ala in Aβ(1–40) increases the helical content and reduces the capacity to form fibrils (Soto et al. 1995). However, although optimal stabilization of helical Aβ by the addition of TFE prevents fibril formation, partial stabilization of helical structure of Aβ instead apparently accelerates formation of β-sheet aggregates and fibrils (Fezoui and Teplow 2002). Furthermore, transient formation of α-helix structure prior to formation of β-sheet structure and fibrils has been observed, suggesting that partially helical forms of Aβ may be on-pathway to fibril formation (Kirkitadze et al. 2001). These studies were performed with Aβ peptides containing the hydrophobic C-terminal part encompassing residues 29–40/42, which forms helical structure in membrane-mimicking solvents, including TFE (Serpell 2000). Our results indicate that stabilization of the Aβ-helix around residues 16–23, as judged by analysis of Aβ(12–28), Aβ(14–23), and Aβ(12–24), reduces aggregation and fibril formation (Figs. 2 ▶, 3 ▶, 5 ▶, 6 ▶, and 8 ▶). Further studies are needed to understand the respective involvement of the central and C-terminal regions of Aβ in fibril formation.

In an unbiased search for sequence determinants of Aβ amyloidogenesis (Wurth et al. 2002), 18 of 36 variants of Aβ(1–42) with reduced tendency to aggregate contained replacements in the region covering residues 17–19. In several cases, the reduction in Aβ aggregation could be rationalized as a result of increased solubility due to the replacements. Of interest in relation to the present investigation, it was found that V18A and F19L mutations reduced aggregation, which would not have been predicted based on effects on solubility only (Wurth et al. 2002). Reduced aggregation as a result of these replacements may be an effect of the increased helical propensity of the Aβ discordant region, similar to the effects observed by V18A/F19A/F20A replacements (Fig. 2 ▶).

The ligand-induced effects indicate that an increase in Aβ discordant helix stability reduces fibril formation. Thus, KAD tripeptide and phospho-L-serine, which increase the helical occupancy of Aβ(12–28) (Figs. 3 ▶, 8 ▶), reduce fibril formation and aggregation (Figs. 4 ▶–8 ▶), whereas structurally related compounds with no effects on secondary structure do not affect fibril formation. In α-helical conformation, the separation of the α-carbons of Lys 16 and Glu 22/Asp 23 is 9–11 Å. In the tripeptide KAD, the positively and negatively charged side chains are separated by ~11 Å, and the charge separation in phospho-L-serine is ~7–8 Å. Thus, dipolar compounds with charge separation that matches the charge separation in the Aβ discordant helix appear to stabilize the helix. However, the lack of effects of phosphoethanolamine and phosphocholine, with similar charge separation as phospho-L-serine, indicate that other factors are also important.

Finally, the effects of phospho-L-serine now observed suggest that membrane phospholipid head-groups can interact with the Aβ discordant helix. Phospholipid membranes have been shown to give variable effects on Aβ fibril formation, with plasma, endosomal, and lysosomal membrane enhancing fibril formation and Golgi membrane preventing it (Waschuk et al. 2001). Modulation of membrane fluidity by cholesterol affects Aβ membrane insertion and fibril formation. At low cholesterol content, Aβ was located in the surface region in mainly β-sheet conformation, whereas at high cholesterol concentrations, Aβ became membrane-inserted in helical conformation and fibril formation was reduced (Ji et al. 2002). It remains to be investigated how membrane phospholipid head-groups interact with Aβ discordant helix, and to what extent such interactions influence the structure of the Aβ region in membrane-associated APP.

Conclusions

Discordant helices, composed of residues with a high β-strand propensity, are found in Aβ, PrP, and other amyloid-forming proteins. In most cases, such helices appear to be stabilized by surrounding protein and lipid environments, but for free Aβ, no such interactions are found. Addition of ligands that stabilize the discordant Aβ α-helix reduces Aβ aggregation and fibril formation in vitro.

Materials and methods

Calculation of accessible surface area

The solvent-accessible surface of each discordant helix residue, in the context of their respective protein, was calculated using the program ICM (Molsoft). The calculations were performed for the same experimentally derived structures as those used for the secondary structure predictions (Kallberg et al. 2001). The relative accessibility for individual residues is derived by dividing their solvent-accessible surface by the standard residue accessibilities calculated for Gly–X–Gly tripeptide segments. A residue is considered buried if the relative accessibility ratio is ≤0.05, and exposed if the relative accessibility ratio is ≥0.3.

Peptides and chemicals

Synthetic peptides corresponding to human Aβ positions 1–40 (amino acid sequence DAEFRHDSGYEVHHQKLVFFAEDVG SNKGAIIGLMVGGVV), 12–28, 12–24, 14–23, and Aβ(12–28) with V18A/F19A/F20A replacements were purchased from Research Genetics or from Interactiva. Aβ(1–40) was purified by reversed-phase HPLC over a C18 column, using a linear gradient of acetonitrile running into 0.1% trifluoroacetic acid for elution. The purified peptide was lyophilized, stored at −20°C, and dissolved shortly before experiments. The tri- and tetrapeptides were synthesized and purified by reversed-phase HPLC (>70% purity) by Interactiva. Phospho-L-serine and phosphoethanolamine were from Fluka, and phosphoglycerol and phosphocholine chloride were from Sigma.

Circular dichroism (CD) spectroscopy

For analysis of secondary structure by CD spectroscopy, Aβ(12–28) or Aβ(12–28; V18A/F19A/F20A) was dissolved at 100 μM concentration in 0%, 30%, 40%, or 70% TFE in 10 mM sodium phosphate buffer (pH 7.0). CD spectra between 180 and 260 nm of Aβ(12–28) or Aβ(12–28; V18A/F19A/F20A) peptides, and of mixtures containing Aβ(12–28) plus 1 mM of tripeptides, phos-pho-L-serine, phosphoethanolamine, phosphocholine, or phosphoglycerol were recorded at 20°C with 2 sec response time, 2 data points/nm, and scan speed 20 nm/min using an AVIV Model 62DS Spectropolarimeter. Spectra of the tripeptides KAD and AAA in 40% aqueous TFE were subtracted from the spectrum of Aβ(12–28) mixed with the corresponding tripeptide. α-Helical contents were calculated from the residual molar ellipticity at 222 nm (Barrow et al. 1992).

Analysis of Aβ fibril formation and aggregation

Fibril formation and aggregation of Aβ(1–40) and fragments covering positions 12–28, 12–24, or 14–23, in the absence and presence of oligopeptides or phospho-compounds were determined. For determination of relative abundance and morphology of fibrils by electron microscopy, the Aβ peptides (100 μM) were incubated for 3 d at 37°C in phosphate-buffered saline (50 mM sodium phosphate/150 mM NaCl at pH 7.4) in the presence or absence of 1 mM of the ligands, and then centrifuged at 20,000g for 20 min. For each experiment, control Aβ samples and those incubated with ligands were divided from the same initial Aβ solution. For analyses of fibrils, the pellets were suspended in a small volume of water by low-energy sonication for 5 sec. Aliquots of 8 μL were placed on electron microscopy grids covered by a formvar film. Excess fluid was withdrawn after 30 sec, and after air-drying, the grids were negatively stained with 2% uranyl acetate in water. The stained grids were examined and photographed in a Philips CM120TWIN electron microscope operated at 80 kV. For an evaluation of the amount of material in the different specimens, the grids (50 mesh) were first scanned at low magnification and the number of larger fibril bundles per grid square counted. The specimens were subsequently examined at high magnification to judge the size of the fibril bundles, the presence of smaller fibril aggregates, and the morphology of the individual fibrils.

Thioflavin T (ThT) fluorescence was used to quantify fibril formation of Aβ(12–28), Aβ(12–28; V18A/F19A/F20F), and of Aβ(12–28) plus 1 mM of ligands after 0, 3, 7, and 14 d incubation. The Aβ peptides were dissolved to 100 μM concentration in 10 mM sodium phosphate buffer (pH 6.0) and incubated at 37°C. At the indicated time points, the peptide solutions were mixed with ThT solution in a 1 : 1 ratio, giving an end concentration of ThT of 10 μM. Fluorescence measurements with excitation wavelength 442 nm and emission at 485 nm were made within 1 min after mixing peptide and ThT.

Aggregation of Aβ(1–40) in the presence of KAD, AAA, or KFFE was studied by determining Aβ contents in 20,000g supernatants at different time points after solubilization. The Aβ(1–40) contents were determined by amino acid analysis of duplicate samples.

Acknowledgments

We are grateful to the Swedish Research Council, the Hans Sigrist Foundation, the Swedish Heart-Lung Foundation, and the King Gustaf V 80th Birthday Fund for support.

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

Abbreviations

  • Aβ, amyloid β-peptide

  • AD, Alzheimer’s disease

  • APP, amyloid precursor protein

  • ASA, accessible surface area

  • CD, circular dichroism

  • PrP, prion protein

  • TFE, trifluoroethanol

  • ThT, thioflavin T

Article and publication are at http://www.proteinscience.org/cgi/doi/10.1110/ps.03442404.

References

  1. Barrow, C.J. and Zagorski, M.G. 1991. Solution structures of β-peptide and its constituent fragments: Relation to amyloid deposition. Science 253 179–182. [DOI] [PubMed] [Google Scholar]
  2. Barrow, C.J., Yasuda, A., Kenny, P.T., and Zagorski, M.G. 1992. Solution conformations and aggregational properties of synthetic amyloid β-peptides of Alzheimer’s disease. Analysis of circular dichroism spectra. J. Mol. Biol. 225 1075–1093. [DOI] [PubMed] [Google Scholar]
  3. Chiti, F., Taddei, N., Baroni, F., Capanni, C., Stefani, M., Ramponi, G., and Dobson, C.M. 2002. Kinetic partitioning of protein folding and aggregation. Nat. Struct. Biol. 9 137–143. [DOI] [PubMed] [Google Scholar]
  4. Coles, M., Bicknell, W., Watson, A., Fairlie, D., and Craik, D. 1998. Solution structure of amyloid β-peptide (1–40) in a water-micelle environment: Is the membrane-spanning domain where we think it is? Biochemistry 37 12700–12706. [DOI] [PubMed] [Google Scholar]
  5. Dima, R.I. and Thirumalai, D. 2002. Exploring the propensities of helices in PrPc to form β sheet using NMR structures and sequence alignments. Biophys. J. 83 1268–1280. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Dobson, C.M. 1999. Protein misfolding, evolution and disease. Trends Biochem. Sci. 24 329–332. [DOI] [PubMed] [Google Scholar]
  7. Esler, W.P. and Wolfe, M.S. 2001. A portrait of Alzheimer secretases—New features and familiar faces. Science 293 1449–1454. [DOI] [PubMed] [Google Scholar]
  8. Fändrich, M., Fletcher, M.A., and Dobson, C.M. 2001. Amyloid fibrils from muscle myoglobin. Even an ordinary globular protein can assume a rogue guise if conditions are right. Nature 410 165–166. [DOI] [PubMed] [Google Scholar]
  9. Fezoui, Y. and Teplow, D.B. 2002. Kinetic studies of amyloid β-protein fibril assembly: Differential effects of α-helix stabilization. J. Biol. Chem. 277 36948–36954. [DOI] [PubMed] [Google Scholar]
  10. Haass, C. and Steiner, H. 2001. Protofibrils, the unifying toxic molecule of neurodegenerative disorders? Nat. Neurosci. 4 859–860. [DOI] [PubMed] [Google Scholar]
  11. Hosia, W., Johansson, J., and Griffiths, W.J. 2002. Hydrogen/deuterium exchange and aggregation of a poly-valine and a poly-leucine α-helix investigated by matrix-assisted laser desorption ionization mass spectrometry. Mol. Cell. Proteomics 1 592–597. [DOI] [PubMed] [Google Scholar]
  12. Hunt, J.F., Earnest, T.N., Bousché, O., Kalghati, K., Reilly, K., Horváth, C., Rothschild, K.J., and Engelman, D.M. 1997. A biophysical study of integral membrane protein folding. Biochemistry 36 15156–15176. [DOI] [PubMed] [Google Scholar]
  13. Janek, K., Rothemund, S., Gast, K., Beyermann, M., Zipper, J., Fabian, H., Bienert, M., and Krause, E. 2001. Study of the conformational transition of Aβ(1–42) using D-amino acid replacement analogues. Biochemistry 40 5457–5463. [DOI] [PubMed] [Google Scholar]
  14. Jarrett, J.T., Berger, E.P., and Lansbury, P.T. 1993. The carboxy terminus of the β amyloid protein is critical for the seeding of amyloid formation: Implications for the pathogenesis of Alzheimer”s disease. Biochemistry 32 4693–4697. [DOI] [PubMed] [Google Scholar]
  15. Ji, S.-R., Wu, Y., and Sui, S.-F. 2002. Cholesterol is an important factor affecting the membrane insertion of β-amyloid peptide (Aβ1–40), which may potentially inhibit the fibril formation. J. Biol. Chem. 277 6273–6279. [DOI] [PubMed] [Google Scholar]
  16. Kallberg, Y., Gustafsson, M., Persson, B., Thyberg, J., and Johansson, J. 2001. Prediction of amyloid fibril-forming proteins. J. Biol. Chem. 276 12945–12950. [DOI] [PubMed] [Google Scholar]
  17. Kelly, J.W. 1998. The alternative conformations of amyloidogenic proteins and their multi-step assembly pathways. Curr. Opin. Struct. Biol. 8 101–106. [DOI] [PubMed] [Google Scholar]
  18. Kirkitadze, M.D., Condron, M.M., and Teplow, D.B. 2001. Identification and characterization of key kinetic intermediates in amyloid β-protein fibrillogenesis. J. Mol. Biol. 312 1103–1119. [DOI] [PubMed] [Google Scholar]
  19. Kirschner, D.A., Inouye, H., Duffy, L.K., Sinclair, A., Lind, M., and Selkoe, D.J. 1987. Synthetic peptide homologous to β protein from Alzheimer disease forms amyloid-like fibrils in vitro. Proc. Natl. Acad. Sci. 84 6953–6957. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Klimov, D.K. and Thirumalai, D. 2003. Dissecting the assembly of Aβ16–22 amyloid peptides into antiparallel β sheets. Structure 11 295–307. [DOI] [PubMed] [Google Scholar]
  21. Näslund, J., Haroutunian, V., Mohs, R., Davis, K.L., Davies, P., Greengard, P., and Buxbaum, J.D. 2000. Correlation between elevated levels of amyloid β-peptide in the brain and cognitive decline. J. Am. Med. Assoc. 283 1571–1577. [DOI] [PubMed] [Google Scholar]
  22. Riek, R., Güntert, P., Döbeli, H., Wipf, B., and Wüthrich, K. 2001. NMR studies in aqueous solution fail to identify significant conformational differences between the monomeric forms of two Alzheimer peptides with widely different plaque-competence, Aβ(1–40)ox and Aβ(1–42)ox. Eur. J. Biochem. 268 5930–5936. [DOI] [PubMed] [Google Scholar]
  23. Selkoe, D.J. 1999. Translating cell biology into therapeutic advances in Alzheimer’s disease. Nature 399 A23–A31. [DOI] [PubMed] [Google Scholar]
  24. ———. 2000. The origins of Alzheimer disease. A is for amyloid. J. Am. Med. Assoc. 283 1615–1617. [DOI] [PubMed] [Google Scholar]
  25. Serpell, L.C. 2000. Alzheimer’s amyloid fibrils: Structure and assembly. Bio-chim. Biophys. Acta 1502 16–30. [DOI] [PubMed] [Google Scholar]
  26. Shao, H., Jao, S., Ma, J., and Zagorski, M. 1999. Solution structures of micelle-bound amyloid β-(1–40) and β-(1–42) peptides of Alzheimer’s disease. J. Mol. Biol. 285 755–773. [DOI] [PubMed] [Google Scholar]
  27. Soto, C., Castano, E., Frangione, B., and Inestrosa, N. 1995. The α-helical to β-sheet transition in the amino-terminal fragment of the amyloid β-peptide modulates amyloid formation. J. Biol. Chem. 271 3963–3967. [DOI] [PubMed] [Google Scholar]
  28. Srisailam, S., Kumar, T.K.S., Rajalingam, D., Kathir, K.M., Sheu, H.-S., Jan, F.-Y., Chao, P.-C., and Yu, C. 2003. Amyloid-like fibril formation in an all β-barrel protein. Partially structured intermediate state(s) is a precursor for fibril formation. J. Biol. Chem. 278 17701–17709. [DOI] [PubMed] [Google Scholar]
  29. Sticht, H., Bayer, P., Willbold, D., Dames, S., Hilbich, C., Beyreuther, K., Frank, R., and Rosch, P. 1995. Structure of amyloid A4(1–40)-peptide of Alzheimer’s disease. Eur. J. Biochem. 233 293–298. [DOI] [PubMed] [Google Scholar]
  30. Thirumalai, D., Klimov, D.K., and Dima, R.I. 2003. Emerging ideas on the molecular basis of protein and peptide aggregation. Curr. Opin. Struct. Biol. 13 146–159. [DOI] [PubMed] [Google Scholar]
  31. Tjernberg, L.O., Näslund, J., Lindqvist, F., Johansson, J., Karlström, A.R., Thyberg, J., Terenius, L., and Nordstedt, C. 1996. Arrest of β-amyloid fibril formation by a pentapeptide ligand. J. Biol. Chem. 271 8545–8548. [DOI] [PubMed] [Google Scholar]
  32. Tjernberg, L.O., Callaway, D.J.E., Tjernberg, A., Hahne, S., Lilliehöök, C., Terenius, L., Thyberg, J., and Nordstedt, C. 1999. A molecular model of Alzheimer amyloid β-peptide fibril formation. J. Biol. Chem. 274 12619–12625. [DOI] [PubMed] [Google Scholar]
  33. Tjernberg, L., Hosia, W., Bark, N., Thyberg, J., and Johansson, J. 2002. Charge attraction and β propensity are necessary for amyloid fibril formation from tetrapeptides. J. Biol. Chem. 277 43243–43246. [DOI] [PubMed] [Google Scholar]
  34. Waschuk, S.A., Elton, E.A., Darabie, A.A., Fraser, P.E., and McLaurin, J. 2001. Cellular membrane composition defines Aβ–lipid interactions. J. Biol. Chem. 276 33561–33568. [DOI] [PubMed] [Google Scholar]
  35. Wurth, C., Guimard, N.K., and Hecht, M.H. 2002. Mutations that reduce aggregation of the Alzheimer’s Aβ42 peptide: An unbiased search for the sequence determinants of Aβ amyloidogenesis. J. Mol. Biol. 319 1279–1290. [DOI] [PubMed] [Google Scholar]

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

RESOURCES