Structure of amyloid β25–35 in lipid environment and cholesterol-dependent membrane pore formation

The amyloid β (Aβ) peptide and its shorter variants, including a highly cytotoxic Aβ25–35 peptide, exert their neurotoxic effect during Alzheimer’s disease by various mechanisms, including cellular membrane permeabilization. The intrinsic polymorphism of Aβ has prevented the identification of the molecular basis of Aβ pore formation by direct structural methods, and computational studies have led to highly divergent pore models. Here, we have employed a set of biophysical techniques to directly monitor Ca2+-transporting Aβ25–35 pores in lipid membranes, to quantitatively characterize pore formation, and to identify the key structural features of the pore. Moreover, the effect of membrane cholesterol on pore formation and the structure of Aβ25–35 has been elucidated. The data suggest that the membrane-embedded peptide forms 6- or 8-stranded β-barrel like structures. The 8-stranded barrels may conduct Ca2+ ions through an inner cavity, whereas the tightly packed 6-stranded barrels need to assemble into supramolecular structures to form a central pore. Cholesterol affects Aβ25–35 pore formation by a dual mechanism, i.e., by direct interaction with the peptide and by affecting membrane structure. Collectively, our data illuminate the molecular basis of Aβ membrane pore formation, which should advance both basic and clinical research on Alzheimer’s disease and membrane-associated pathologies in general.

10 (e,f), 0.20 (g,h), and 0.40 (i,j). At each xchol, spectra were measured at ǁ and  polarizations, as indicated by labels "parallel" and "perpendicular." The measured spectra, shown in red dotted lines, were curve-fitted to generate the amide I components that are shown under the spectra as follows: turns-black dotted lines, helix-green solid lines, irregular structure-gray solid lines, -sheet-red solid lines, and side chains-black dashed lines. In panels (a) and (b), turn, -helix, irregular structure, and -sheet are labeled t, , , and , respectively. The curvefit, i.e. the sum of all components, is shown for each spectrum in solid blue line. 10 (e,f), 0.20 (g,h), and 0.40 (i,j). Spectra were measured at ǁ and  polarizations, as indicated by labels "parallel" and "perpendicular." All other details are as in Fig. S4.

Materials
The synthetic A25 -35 peptide, acetylated at N-terminus and amidated at C-terminus, was purchased from Peptide 2.0 Inc. (Chantilly, VA, USA) and was 98% pure, as verified by high performance liquid chromatography and mass-spectrometry. Quin-2 tetrapotassium salt was from EMD Chemicals (San Diego, CA). Hexafluoroisopropanol (HFIP), non-fluorescent Ca 2+ ionophore 4-Br-A23187, salts, buffers, and most of other reagents were purchased from Sigma-Aldrich ( The solvent was removed by desiccation, the dry lipid mixture was suspended in an aqueous buffer containing 75 mM KCl, 6 mM Quin-2 tetrapotassium salt, 20 mM Tris-HCl (pH 7.2), and extruded through 100 nm pore-size polycarbonate filters using a mini extruder (Avanti Polar Lipids). The lipid suspension was passed through a column loaded with Sephadex G-50 resin, equilibrated with 75 mM NaCl, 30 mM myo-inositol, 20 mM Tris-HCl (pH 7.2), to remove external Quin-2. Elution samples containing the highest lipid content, assessed by apparent turbidity, were collected and the total lipid concentration was adjusted to 0.2 mM based on a calibration curve, i.e. dependence of right-angle static light scattering on lipid concentration. Adjustment of the lipid concentration was paralleled with addition of 6 mM CaCl2 to the vesicles. Care was taken to maintain the osmotic balance, using myo-inositol when necessary. Thus, intravesicular Quin-2 was sequestered from external Ca 2+ ions.
For membrane permeabilization experiments, a certain amount of HFIP solution of the A25-35 peptide was dried by desiccation, then aqueous buffer (75 mM NaCl, 20 mM Tris-HCl, pH 7.2) was added and the sample was stirred in a glass vial for 2.5 hours. To monitor peptideinduced membrane permeabilization, a certain volume of Quin-2 loaded vesicles was placed in a quartz cuvette with 4 mm  4 mm internal cross section and the baseline fluorescence of Quin-2, entrapped in lipid vesicles, was measured at excitation wavelength of 339 nm and emission wavelengths between 450 and 600 nm, using a J-810 spectropolarimeter with a fluorescence attachment (Jasco, Tokyo, Japan). Upon establishment of a stable baseline, a certain dose of the peptide, pre-incubated in aqueous buffer as described above, was added and consecutive fluorescence spectra were collected for 15 minutes until the increase in fluorescence approached saturation. In positive control experiments, non-fluorescent Ca 2+ ionophore 4-Br-A23187 was added, which produced the maximum increase in Quin-2 fluorescence, Fmax. In negative control experiments, blank buffer was added, which only resulted in slight decrease in fluorescence due to sample dilution. The rate constant of the exponential increase in Quin-2 fluorescence and the level of saturation were used to analyze the pore formation kinetics, as described in section Methods of the main text.

Circular dichroism and light scattering.
At the end of fluorescence measurements, circular dichroism (CD) and right-angle static light scattering spectra of the vesicle samples were measured, using the same J-810 spectropolarimeter, to estimate the secondary structure of the peptide in the presence of vesicles and to verify the integrity of the vesicles. CD was measured between 180 and 320 nm. The spectra were normalized to obtain the mean residue molar ellipticity, [] = meas/cnresl, where meas is the measured ellipticity in millidegrees, c is the molar concentration of the peptide, nres is the number of amino acid residues in the peptide, and l is the optical path-length in millimeters. Light scattering was measured in the fluorescence mode, i.e., using the fluorescence photomultiplier mounted at 90 degrees relative to the incident beam, using incident wavelength at 550 nm and monitoring the intensity of scattered light between 535 and 565 nm. The signal intensity was obtained by subtracting the baseline value from the maximum.
Membrane fluidity. Membrane fluidity was measured by the method of Laurdan generalized polarization (GP). Laurdan was incorporated in vesicle membranes at 1 mol % and fluorescence spectra were measured between 380 and 580 nm with excitation at 360 nm. GP of Laurdan was evaluated as GP = (F435 -F500)/(F435 + F500), where F is the fluorescence emission intensity at respective wavelength.

Polarized ATR-FTIR experiments.
Lipid multilayers with incorporated A25-35 peptide were prepared as follows. HFIP solution of the peptide was mixed with chloroform solutions of POPG, POPC, and cholesterol at molar proportions of POPG/POPC/cholesterol = 0.3/(0.7xchol)/xchol and a peptide/(total lipid) molar ratio of 1:15. As in membrane leakage experiment, lipid layers with the following cholesterol fractions were studied: xchol = 0.00, 0.05, 0.10, 0.20, and 0.40. The lipid-peptide solution was carefully and uniformly spread over a germanium plate (5.0 cm  2.0 cm  0.1 cm, cut at the 2.0 cm sides at a 45 o aperture angle), which served as an internal reflection element in attenuated total reflection Fourier transform infrared (ATR-FTIR) experiments. The sample was air-dried, followed by desiccation for 1 hour. The plate, with the dry peptide-lipid sample deposited on one side, was assembled into a perfusable sample holder, which was mounted on a four-mirror ATR system (Buck Scientific, East Norwalk, CT, USA) and placed in a Vector 22 FTIR spectrometer (Bruker, Billerica, MA, USA), equipped with a liquid nitrogen-cooled Hg/Cd/Te detector and an aluminum-grid-on-KRS-5 polarizer (Specac, Newmarket, Suffolk, UK). The spectrometer was purged with dry air (passed through a silica-gel column) for 15 minutes, then transmission spectra were collected at 2 cm -1 nominal resolution at parallel and perpendicular (ǁ and  ) polarizations, i.e., with plane-polarized incident light when the electric vector of the radiation is oriented parallel or perpendicular to the plane of incidence, respectively. Following completion of the measurements of the spectra of dry sample, the cell was disassembled and the plate with the lipid-peptide sample was exposed to D2O vapors by incubation with hot (~90 o C) D2O in a closed glass chamber for 15 minutes. Appearance of dew on the internal surface of the chamber indicated saturation of the internal volume with D2O vapor. The sample, hydrated by D2O from gas phase, was assembled in the cell and FTIR spectra were recorded at ǁ and  polarizations. Finally, a buffer composed of 50 mM NaCl and 50 mM Na,K-phosphate in D2O (pH* = 6.8) was injected into the cell and the spectra of the sample exposed to bulk buffer were measured again at both polarizations. (pH* is the pH-meter reading and corresponds to pD = pH* + 0.4 = 7.2) 1,2 . In separate experiments, the transmission spectra of the bare germanium plate and those of identical lipid multilayers without the peptide were measured at ǁ and  polarizations, which were used as reference for calculation of the absorbance spectra.

Data analysis procedures
The oligomeric state of the pore. The oligomeric state of the pore has been evaluated based on a formalism of reversible aggregation, described in detail by Garg et al. 3 and in earlier publications cited therein. Values of relative equilibrium levels of Quin-2 fluorescence, Frel = Feq/Fmax, along with peptide-peptide affinity constants (Kp) and membrane-bound peptide concentrations ([Pb]), were used to quantitate the number of peptide oligomers involved in pore formation, n, based on the following relationship 3 : Peptide structure from ATR-FTIR spectroscopy. The secondary structure of the peptide embedded in lipid layers was determined by analysis of the ATR-FTIR amide I absorbance band. The second derivatives of ATR-FTIR spectra measured at ǁ and  polarizations were used to identify the number and the locations of amide I components. Curve-fitting procedures were then carried out using the GRAMS software, which produced the actual amide I components. These components were assigned to certain secondary structures based on the following spectral ranges for each secondary structure: various types of turns (1700-1665 cm -1 ), -helix (1664-1648 cm -1 ), irregular structure (1647 and 1639 cm -1 ), -sheet (1638-1620 cm -1 ) 4 . The low frequency amide I components (1619-1600 cm -1 ) were assigned to side chains and were subtracted from the amide I bands in secondary structure assessment procedures. Thus, the areas of components assigned to defined secondary structures were obtained for ǁ and  polarizations of the incident light: at,ǁ, a ,ǁ , a ,ǁ , a ,ǁ , at,  , a , , a , , and a , , where the subscripts t, ,  and  stand for turn, -helix, irregular, and -sheet structures, respectively. The relative areas of amide I components measured at a certain polarization cannot be directly used to evaluate the fractions of respective secondary structures because absorbance intensities depend not only on the content but also on the orientation of each structure. To identify the fraction of each secondary structure, first the genuine (polarization-independent) relative amide I areas were determined: The following extinction coefficients have been used:   = 5.110 7 cm/mol,   = 7.010 7 cm/mol, t =5.510 7 cm/mol, and   = 4.510 7 cm/mol 4 .
Peptide orientation from ATR-FTIR spectroscopy. The ATR dichroic ratios for each secondary structure were calculated as Ri = ai,ǁ/ai,  . The orientational order parameter, S, of a structure that has a molecular axis, such as an -helix or an acyl chain of a lipid molecule, is defined as 4 :   In Eq. S4,  is the angle between the transition dipole moment and the molecular axis, 2 2 2 x y z B E RE E    , Ex, Ey, and Ez are the orthogonal components of the incident infrared radiation, R is the ATR dichroic ratio, and the angular bracket indicates average value. Under our experimental conditions (a lipid-peptide multilayer on a germanium plate, with refractive indices of 1.43 and 4.00, respectively), Ex = 1.399, Ey = 1.514, and Ez = 1.621. For an -helix, the angle between the transition diploe moment of amide I vibrational mode and helical axis is  = 391degrees, and for a lipid acyl chain in all-trans conformation, the angle between CH2 stretching vibrations and the chain axis is  = 90 degrees 4-6 . The order parameters for -helices