Eur. J. Biochem. 269, 2238–2246 (2002)
FEBS 2002

doi:10.1046/j.1432-1033.2002.02882.x

The amyloid precursor protein interacts with neutral lipids Liposomes and monolayer studies Raghda Lahdo1, Ste´phane Coillet-Matillon1, Jean-Paul Chauvet2 and Laurence de La Fournie`re-Bessueille1 1

Laboratoire de Physico-Chimie Biologique, Universite´ Claude Bernard, Lyon, France; 2IFOS, Equipe Bioinge´nierie et Reconnaissance Ge´ne´tique, Ecully, France

The amyloid protein precursor (APP) was incorporated into liposomes or phospholipid monolayers. APP insertion into liposomes required neutral lipids, such as L-a-phosphatidylcholine, in the target membrane. It was prevented in vesicles containing L-a-phosphatidylserine. The insertion was enhanced in acidic solutions, suggesting that it is modulated by specific charge/charge interactions. Surfaceactive properties and behaviour of APP were characterized during insertion of the protein in monomolecular films of L-a-phosphatidylcholine, L-a-phosphatidylethanolamine or L-a-phosphatidylserine. The presence of the lipid film enhanced the rate of adsorption of the protein at the interface, and the increase in surface pressure was consistent with APP penetrating the lipid film. The adsorption of APP on the lipid monolayers displayed a significant head group dependency, suggesting that the changes in surface pressure

produced by the protein were probably affected by electrostatic interactions with the lipid layers. Our results indicate that the penetration of the protein into the lipid monolayer is also influenced by the hydrophobic interactions between APP and the lipid. CD spectra showed that a large proportion of the a-helical secondary structure of APP remained preserved over the pH or ionic strength ranges used. Our findings suggest that APP/membrane interactions are mediated by the lipid composition and depend on both electrostatic and hydrophobic effects, and that the variations observed are not due to major secondary structural changes in APP. These observations may be related to the partitioning of APP into membrane microdomains.

Limited proteolysis of the amyloid precursor protein (APP) generates the amyloid-b-protein (Ab), which is a major component of brain senile plaques in Alzheimer’s disease (AD) [1,2]. APP occurs in neural and non-neural tissues as several membrane-associated glycoproteins of 110–135 kDa [3]. It is a N- and O-glycosylated single-chain molecule consisting of 770 amino acid residues, with an isoelectric point of 4–5. The APP gene is expressed in brain and in several peripheral tissues, but the physiological functions of APP and its role in the disease are still poorly understood. A recent report proposed that APP normally behaves in the brain as a cell surface signalling molecule, and that an alteration of this function is one of the possible causes of the neurodegeneration and consequent dementia in AD [4]. The Ab peptide is produced in the endosomal compartment and in the endoplasmic reticulum or Golgi complex [5,6] through the sequential action of b- and c-secretases [7–12]. APP could also be cleaved by an a-secretase, within the Ab

sequence, thus preventing amyloidogenesis, and results in the secretion of the larger soluble amino-terminal product (sAPP; reviewed [13]). The molecular mechanism involved in APP cleavage and Ab production has still to be resolved. Minor changes of the membrane lipid composition could affect the stability of APP as well as its processing, or alter the function of secretases within the membrane and their activities towards APP. For example, modifications of the cholesterol content result in the alteration of Ab secretion [14], or in the variation of sAPP release from neuronal cells [15]. Recent papers suggest that cholesterol levels regulate Ab production and Alzheimer’s disease pathology by acting on the multiple enzymes which regulate the APP processing [16–18]. X-ray diffraction studies show that membranes isolated from AD brains are thinner than those obtained from age-matched control brains [19]. This alteration in membrane thickness may change the spatial relationship between membrane-associated proteases and APP, modifying the amyloidogenic cleavage of the protein. In vivo, monomeric Ab appears to be a normal constituent of cerebrospinal fluid but during normal aging or as a result of a disease process, Ab self associates into fibres that precipitate as plaques in the brain. In vitro, the toxicity of Ab is clearly correlated with the aggregation into crossb-pleated sheet fibrils. This process is enhanced by increasing the concentration of Ab or by altering the pH or the ionic strength [20,21]. The fibrillogenic properties of the Ab peptide are highly dependent on the membrane composition [22–25]. A growing number of studies indicate that Ab may alter the physicochemical properties of neuronal membranes, including membrane fluidity, membrane permeability to ions and lipid peroxidation [26–29]. These studies

Correspondence to L. de La Fournie`re – Bessueille, Laboratoire de Physico-Chimie Biologique, UMR CNRS 5013, Baˆtiment E. Chevreul, Universite´ Claude Bernard, Lyon I, 43 bd du 11 novembre 1918, 69622 Villeurbanne cedex, France. Fax: + 33 4 72 43 15 43, Tel.: + 33 4 72 44 83 24, E-mail: [email protected] Abbreviations: AD, Alzheimer’s disease; Ab, amyloid peptide; APP, amyloid precursor protein; b-OG, n-octyl b-D-glucopyranoside; LUV, large unilamellar vesicles; PtdCho, phosphatidylcholine; PtdEtn, phosphatidylethanolamine; PtdSer, phosphatidylserine. (Received 10 January 2002, revised 11 March 2002, accepted 15 March 2002)

Keywords: amyloid precursor protein; liposomes; monolayer; phospholipids; protein–lipid interactions.


FEBS 2002

Interaction of amyloid precursor protein with lipids (Eur. J. Biochem. 269) 2239

show that at least some of the physiological effects of Ab may involve direct interactions between the peptide and the membrane lipids. However, the mechanisms of Ab–membrane interaction are still unclear. As no data exist on the binding of APP itself to the lipid membrane, it is of great interest to examine the interactions of APP with lipids. In the present study, we have investigated the ability of APP to interact with lipids. The potential interactions of the protein with lipid membranes were characterized using two approaches. In the first, the insertion of APP into liposomes was analysed via a detergent-mediated reconstitution procedure. In the second, the interaction of the protein with a monomolecular lipid film was characterized using the Langmuir technique. The role of electrostatic and hydrophobic effects in APP association with lipids was investigated in more detail under various conditions involving different ionic strengths and pH, with respect to the structure of the protein.

mixed with an equal volume of 60% (w/v) sucrose in 20 mM Tris/HCl, 150 mM NaCl, pH 7.4 buffer. A 5–25% sucrose gradient was poured on the top of the 30% layer and samples were centrifuged at 160 000 g for 3 h at 4 C. Fractions (0.5 mL) from the gradient were subsequently collected from the bottom of the tube and the lipid and protein contents were analysed. The lipid content was determined by comparison with a specific experiment using 3 H[PtdCho] detection. Lipids were found to be present mainly at a sucrose concentration of
15%. The presence of APP was determined by measurement of 125I-labeled APP radioactivity. Percentage incorporation was defined as the ratio between the radioactivity recovered in liposomal fractions and the total radioactivity loaded on the gradient. It corresponds to the percentage of APP inserted into liposomes. The reconstitution was performed under different conditions, at varying pH and salt concentrations. Monolayer measurements

MATERIALS AND METHODS APP was purified from porcine brains as described previously [30]. The procedure yields homogeneous preparations as judged by SDS/PAGE. The protein concentration was determined by the Bradford assay [31] using BSA as a standard. L-a-phosphatidylcholine from egg yolk (PtdCho), L-a-phosphatidylethanolamine from egg yolk (PtdEtn), L-a-phosphatidylserine from bovine brain (PtdSer) and n-octyl-b-D-glucopyranoside (b-OG) were from Sigma. 125 I-labeled APP was prepared by the chloramine T method to obtain a final specific radioactivity of 79 MBqÆnmol)1. All other reagents were of analytical grade. Preparation of liposomes Vesicles were obtained by dialysis as described by Angrand et al. [32] or by extrusion as follows: large unilamellar vesicles (LUV) were prepared from a phospholipid stock solution dissolved in chloroform/methanol (2 : 1). The resulting solution was then evaporated to dryness under a stream of nitrogen and the last traces of solvent subsequently removed by a further 3–6 h evaporation period under vacuum. The remaining lipid film was then hydrated in 20 mM Tris/HCl, 150 mM NaCl pH 7.4 at a higher temperature (room temperature) than the phase transition temperature of the corresponding lipid and dispersed vigorously by vortexing. LUV were formed using six fast freeze–thaw cycles. They were subsequently extruded 19 times through two polycarbonate membranes (400 and 200 nm pore size), using a mini extruder (Avanti Polar). The final phospholipid concentration was 20 mgÆmL)1.

Measurements were recorded at 21 C using a Teflon trough (Riegler and Kirstein, Germany). Adsorption of the protein was performed at a constant surface area either in a small Teflon dish (diameter, 1.6 cm) with a subphase volume of 4 mL or with a larger Teflon dish (diameter, 3.4 cm) with a subphase volume of 19 mL. The surface pressure was measured as a function of time by the Wilhelmy plate method using plates cut from filter paper (Whatman no. 1) and a computer-controlled transducer readout. The surface activity of the protein was recorded at a free water interface and characterized in the presence of a preformed lipid monolayer. Lipid monolayers were spread at the air–water interface from a chloroform solution to give an initial surface pressure (pi). Ten minutes after the formation of the monolayer, a desired volume of APP in 20 mM Tris/HCl, 150 mM NaCl pH 7.4 or in Tris/maleate 20 mM, 150 mM NaCl pH 7, was injected below the surface. The subphase was stirred continuously with a Teflon-coated stirring bar and a magnetic stirrer. In similar experiments, we examined the effect of pH or NaCl concentrations in the subphase buffer on the adsorption characteristics of APP in the absence or in the presence of a phospholipid monolayer. CD spectroscopy Far-UV CD spectra of APP were recorded on a Jobin-Yvon CD6 spectropolarimeter. All measurements were done at 25 C in a 0.05-mm path length quartz cell with an APP concentration of 0.15 mgÆmL)1 in the appropriate buffer (10 mM NaH2PO4/Na2HPO4). Blanks (various buffers with or without lipids) were routinely recorded and substracted from the original spectra.

Preparation of phospholipid–protein complexes The incorporation of APP into liposomes was performed using the detergent-mediated procedure described by Le´vy et al. [33]. Briefly, the complexes of phospholipid and protein were prepared at a phospholipid/protein ratio of 500 : 1 (w/w). The lipid/protein mixture containing trace amounts of 125I-labeled APP was incubated with the desired concentration of b-OG. The excess of detergent was then removed by extensive dialysis against 20 mM Tris/HCl, 150 mM NaCl, pH 7.4 buffer. The resulting solution was

RESULTS Incorporation of APP into liposomes We studied the ability of the protein to insert into preformed liposomes, via a detergent-mediated procedure, to investigate the potential association of APP with a lipid membrane. The complexes generated between the protein and lipids in the presence of b-OG were fractionated on a discontinuous sucrose density gradient and the amount of


FEBS 2002

2240 R. Lahdo et al. (Eur. J. Biochem. 269)

Fig. 1. Density gradient centrifugation profiles for APP liposomes reconstituted from b-OG. Liposomes (1 mg) were incubated with b-OG (16 mM) and APP. The detergent was removed by dialysis, the samples were then submitted to a flotation on discontinuous sucrose gradients. Fractions (0.5 mL) were collected from the bottom of the gradient and APP content was measured by assaying for 125I-labeled APP. (- - -, 1 lg APP; –––, 2 lg APP; – ) –, 5 lg APP).

radiolabelled APP associated with liposomes was determined. Control experiments showed that all the radioactivity associated with pure APP in the absence of liposomes was recovered at the bottom of the gradient (25% sucrose), while pure phospholipid liposomes were collected at a sucrose concentration of 15%. After a 20-h dialysis period of the reconstitution mixture composed of APP, PtdCho liposomes and b-OG, we observed a comigration of phospholipid and protein over a relatively narrow density range, at
15% sucrose. A small fraction of the protein was associated with the lipid, and
80% free protein was recovered at the bottom of the gradient (Fig. 1). The presence of the free protein indicates a low incorporation efficiency of APP during the reconstitution procedure. The influence of lipid/protein ratio on efficiency of protein incorporation into liposomes during reconstitution was examined. Proteoliposomes samples containing APP and PtdCho liposomes were incubated in the presence of b-OG at initial phospholipid/protein ratios of 1000 : 1, 500 : 1 or 200 : 1 (w/w). Most of the protein migrated at the bottom of the gradient (Fig. 1), which indicated incomplete incorporation of APP into liposomes whatever the lipid/protein ratio. Therefore, the following experiments were carried out with phospholipid/protein ratio of 500 : 1. Reconstitution experiments performed with ionic or zwitterionic detergents were less efficient and less reproducible. The role of different phospholipids in the insertion of APP into lipid bilayers was examined (Fig. 2). Vesicles containing PtdEtn/PtdCho (molar ratio 62.5 : 37.5) did not change the incorporation rate of APP into the membrane. However, PtdSer-containing liposomes totally prevented the binding of APP to the phospholipid membrane. These results suggest that charge– charge interactions between phospholipids and APP are involved in the protein insertion process to the membrane. We repeated these experiments at various pH or with higher concentrations of NaCl in the buffer. The APP incorporation was increased threefold to fourfold at acidic pH for neutral phospholipids (PtdCho or PtdCho/PtdEtn vesicles)

Fig. 2. Effect of lipid composition and pH on the incorporation of APP into phospholipid vesicles. Percentage of incorporation of APP into liposomes composed of 1 mg phospholipids (d, PtdCho; r, PtdCho/ PtdEtn; j, PtdSer) mixed with b-OG (16 mM, 20 mM and 14 mM, respectively) and APP (2 lg) at different pH. After dialysis against the same buffer as that used for incubation, proteoliposomes were separated on a discontinuous sucrose gradient. The 125I-labeled APP was measured in the collected fractions (0.5 mL). The percentage of incorporation was calculated as follows: [125I-labeled APP radioactivity associated with liposomes/total 125I-labeled APP radioactivity loaded on the gradient] · 100.

as shown in Fig. 2. These results indicate that APP insertion in the lipid bilayer is enhanced at a pH close to the pI of the protein. An increasing salt concentration lead to a decreased incorporation rate of APP into preformed LUVs to approximatively zero at 0.3 M NaCl (data not shown). These data show that APP insertion in the lipid bilayer is inhibited by a high concentration of salt. Monolayer experiments Adsorption of APP at the air–water interface. The surface pressure (p) was measured as a function of time for various subphase concentrations of APP in the absence of preformed lipid films (subphase: 20 mM Tris/HCl, 150 mM NaCl, pH 7.4) (Fig. 3). When the subphase concentration of APP was < 0.2 lgÆmL)1, no significant changes in the surface pressure were detected during 120 min (not shown). The effects of higher concentrations of APP on the surface pressure are shown in Fig. 3A. The results showed that the final surface pressure increased with APP concentration in the subphase, an indication that the interface was not saturated by the protein, for concentrations up to 1 lgÆmL)1. Moreover, the surface pressure increase does not occur immediately after the injection of the protein into the subphase (Fig. 3A). For protein concentrations of 0.35, 0.5, 1 and 2 lgÆmL)1 the surface pressure started to increase after 90, 50, 20 and 10 min, respectively. For the lowest APP concentration used, the observed lag time reached several hours. The sigmoidal shape of the p vs. time curve suggests the occurrence of a co-operative process. These results indicate that the adsorption of APP at the air–water interface is both concentration- and time-dependent. The process of the protein adsorption can be analysed by a first order equation:


FEBS 2002

Interaction of amyloid precursor protein with lipids (Eur. J. Biochem. 269) 2241

perhaps unfolding at the surface layer, while s2, defined by the second linear part, can be related to a subsequent interfacial rearrangement of the adsorbed protein molecules. When APP concentrations were 0.35 and 1 lgÆmL)1, s1 reached 200 and 23 min, while s2 was around 25 and 75 min, respectively. From these results, s1 and s2 appear to be very sensitive to the bulk concentration of APP. In the absence of a lipid monolayer, APP adsorbed at the air–water interface, leading to a stable monolayer. The surface activity of APP resulted in a protein monolayer with a surface pressure of
12 mNÆm)1 at pH 7 on a buffer subphase containing at least 150 mM of NaCl (Fig. 3B). The presence of 300 mM NaCl in the subphase buffer increased the surface activity of APP to
14 mNÆm)1 and significantly decreased the lag time (Fig. 3B). This result may indicate an increase of the hydrophobicity of the protein due to charge neutralization. The presence of 5 mM NaCl in the subphase buffer decreased the surface activity to 4.5 mNÆm)1 with a 2-h lag time (Fig. 3B). The surface activity of the protein in the absence of lipids was also studied on subphase buffers of different pH (Fig. 3C). Lowering the pH to 6 did not promote changes on the final surface pressure of the protein monolayer, but increased the lag time threefold. At a pH below 6, longer lag times and lower surface activities for APP were observed (pe ¼ 10 mNÆm)1 at pH 5, and pe ¼ 2 mNÆm)1 at pH 4) (Fig. 3C). This result may be explained by the progressive reduction of the global charge of the protein as the pH approaches the isoelectric point, giving a less hydrophilic character to the APP molecule.

Fig. 3. Kinetics recording the surface behaviour of APP at the air–water interface. (A) p–t curves corresponding to the penetration of APP into the air–water interface under different bulk concentrations of the protein (buffer: Tris/HCl 20 mM, NaCl 150 mM, pH 7; Teflon trough 19 mL, 9 cm2). Insert: Ln (1-p/pe)) vs. t plots for the adsorption of APP into the air–water interface. Protein concentrations in the bulk are 0.35 lgÆmL)1 (a) and 1 lgÆmL)1 (b). (B) p–t curves corresponding to the penetration of APP into the air–water interface in the presence of various salt concentrations (buffer: Tris/maleate 20 mM, pH 7; Teflon trough 4 mL, 2 cm2). (C) p–t curves corresponding to the penetration of APP into the air–water interface under various pH (buffer: Tris/maleate 20 mM, NaCl 150 mM; Teflon trough 4 mL, 2 cm2).

 ln

pe  p pe  p0

 ¼ 

t s

ðEqn 1Þ

where pe, p, p0, are the surface pressure values at steadystate conditions, at time t and at time t ¼ 0, respectively, and s is the relaxation time. In order to evaluate the parameters that control the successive steps of the adsorption of APP at the air/water interface, the plots of ln(1–p/pe) vs. time were obtained according to Eqn 1. The curves (insert in Fig. 3A) present at least two linear parts. According to Graham & Phillips’ work [34], the relaxation time s1, corresponding to the first linear part, is corelated with the protein adsorption and

Interactions of APP with phospholipid monolayers. Increase in surface pressure (Dp) vs. time curves (Fig. 4A) observed for APP adsorption into films of PtdCho (initial pi ¼ 10 mNÆm)1), at different subphase concentrations, demonstrated that the rate of adsorption of the protein at the lipid–water interface increased considerably as compared with the surface activity of the protein at the air–water interface (i.e., the time to obtain the maximum Dp was decreased, even at low protein concentrations). The specific protein concentration used induced no change in surface tension of the buffer–air interface during 1 h, after which the protein was injected into the subphase in the absence of preformed phospholipid monolayer (Fig. 3A). However, the increase of surface pressure of a preformed phospholipid monolayer occurred immediately after injection of the protein under the lipid layer. These results suggested a very fast diffusion and incorporation process of APP into the lipid film. The presence of the PtdCho monolayer slightly increased the final value of the surface pressure increase produced by the protein at this relatively low initial surface pressure (Fig. 4A). The effect of the ionic strength was studied in order to understand the role of electrostatic shielding in the binding of APP to the phospholipid membrane. Qualitatively, high ionic strength was found to increase the binding of APP to the lipid membrane (Fig. 4B). The effect of pH on the APP surface activity was also analysed. In contrast to the pH dependence observed with APP during the detergent-mediated reconstitution, no changes in the surface pressure due to the incorporation of the protein was observed in a pH range of 4–8 (Fig. 4C). Finally, the interactions of APP with spread monolayers of various phospholipids were studied at a low concentration of APP. Fig. 5A shows typical records of the


FEBS 2002

2242 R. Lahdo et al. (Eur. J. Biochem. 269)

Fig. 4. Kinetics recording the surface behaviour of APP at the phosphatidylcholine–water interface. (A) p–t curves of APP inserted into the PtdCho monolayer under different bulk concentrations of the protein (buffer: Tris/HCl 20 mM, NaCl 150 mM, pH 7; Teflon trough 19 mL, 9 cm2). (B) p–t curves of APP inserted into the PtdCho monolayer under different salt concentration (buffer: Tris/maleate 20 mM, pH 7; Teflon trough 4 mL, 2 cm2). (C) p–t curves of APP inserted into the PtdCho monolayer under different pH (buffer: Tris/maleate 20 mM, NaCl 150 mM; Teflon trough 4 mL, 2 cm2).

surface pressure as a function of time corresponding to the insertion of APP into PtdCho, PtdEtn or PtdSer monolayers spread at an initial surface pressure of pi ¼ 10 mNÆm )1± 1 mNÆm)1. PtdCho, PtdEtn and PtdSer showed essentially identical surface pressure-area isotherms and thus have an identical head group spacing at a given surface pressure, despite some potential differences in acyl group conformations [35]. The adsorption of the protein to the phospholipid monolayer produced a change in the surface pressure (Dp) of
5 mNÆm)1 for both PtdCho and PtdEtn monolayers (Fig. 5A). The same surface pressure increase was observed when APP was diluted (final concentration of 0.35 lgÆmL)1) in the buffer subphase before the lipid monolayer was formed (data not shown). Comparison of the results for pi ¼ 5 mNÆm)1 with those for higher pi are illustrated in Fig. 5B [Dp ¼ f(pi)]. The data suggest that the increase in the surface pressure Dp due to

Fig. 5. Surface pressure increase after protein injection under different phospholipid monolayers. (A) Penetration pressure (Dp) induced by the insertion of APP into the phospholipid monolayers. (- - -, PtdCho; —–, PtdEtn; – ) –, PtdSer). Buffer: Tris/HCl 20 mM, NaCl 150 mM, pH 7. The initial surface pressure was 10 ± 1 mNÆm)1 (Teflon trough 19 mL, 9 cm2). (B) Surface pressure increase after protein injection with respect to the initial surface pressure of the phospholipid film. (d, PtdCho; r, PtdEtn: j, PtdSer). Buffer: Tris/HCl 20 mM, NaCl 150 mM, pH 7. (Teflon trough 19 mL, 9 cm2).

the adsorption of APP on the phospholipid monolayers is dependent on the initial surface pressure of the lipid film. The similarity of the Dp vs. time profiles for the phospholipids PtdCho and PtdEtn (Fig. 5A) confirmed the observation that APP did preferentially interact with either one of the two phospholipids during adsorption at the air–water interface. When APP was injected below monolayers of PtdCho, PtdEtn or PtdSer formed at pi above 23, 19 and 11 mNÆm)1, respectively, the protein was no longer able to induce any increase of the surface pressure (Fig. 5B). These critical surface pressures for APP penetration (pc) correspond to the extrapolated initial surface pressures beyond which no increase in the surface pressure occurred. CD spectroscopy NaCl or pH effect on the secondary structure of APP. As the interaction of APP with lipids is either pH or NaCl dependent, we have considered the possibility that these variations may lead to protein conformational changes. The effect of NaCl on the secondary structure of APP was studied in the far UV region. Fig. 6A shows the CD spectra


FEBS 2002

Interaction of amyloid precursor protein with lipids (Eur. J. Biochem. 269) 2243

Fig. 6. CD of APP. (A) CD spectra of APP in phosphate buffer 10 mM pH 7 as a function of salt concentration (—–, No NaCl; – ) – 150 mM NaCl; - - -, 500 mM NaCl). (B) CD spectra of APP in phosphate buffer 10 mM NaCl 150 mM as a function of pH (—–, pH 7; –– –– , pH 6, – ) –, pH 5; - - -, pH 4).

of APP in 10 mM NaH2PO4/Na2HPO4 buffer pH 7 at varying concentrations of NaCl. The CD spectra were smoothed, and an analysis of the CD spectrum of APP at neutral pH and ionic strength (Fig. 6A) yielded 40% a helix and 20% b sheet structures as reported previously [30]. No significant variation was observed when the NaCl concentration increased. A slight increase (10%) of the a helix content was observed for a NaCl concentration of 0.5 M. Far-UV CD spectra indicated that APP remains mainly a-helical at mildly acidic pH, reflecting no major change in the secondary structure between pH 5 and 7 (Fig. 6B). The spectrum of APP at pH 4 is somewhat flatter than the others, reflecting a possible aggregation of the protein as a consequence of its denaturation in the acidic buffer. Effect of LUVs. The protein conformation in the presence of neutral vesicles was analysed by CD spectroscopy (data not shown). A small decrease of the a helix content of the protein was observed in the presence of 2 mM PtdCho vesicles. Neutral phospholipid vesicles had a very weak influence on the protein conformation.

DISCUSSION In order to get insight into the role of lipids in APP processing we have characterized its association with artificial lipid vesicles or lipid monolayers. APP was able to interact with the membranes of liposomes, although

incomplete APP incorporation into liposomes could be observed with b-OG-mediated reconstitution. The nonionic detergent b-OG was used because it is a nondenaturating detergent with a high critical micellar concentration (20–25 mM). Therefore, it can be rapidly removed by dialysis [33,36]. The low incorporation rates observed in b-OG-mediated reconstitution were, however, in the same range as those observed by Lin et al. [26] for Ab incorporated into liposomes using the sonification technique. The ability of APP to insert into neutral lipid vesicles was dependent on pH. Indeed, the binding was greater at low pH, when the net charge of the protein is close to a nil value, than at neutral pH, where APP is in an anionic form. It is more likely that a low pH is required for the protonation of the negatively charged carboxyl groups of aspartic or glutamic acids, whose pK values are near 4. The protonation of the carboxyl functions reduce significantly the repulsion between the negatively charged groups on the membrane and the negatively charged amino acids present on APP. This requirement for a low pH suggests that electrostatic repulsion prevents the protein–membrane association. Furthermore, no incorporation was obtained in the presence of negatively charged lipids in the membrane, whatever was the pH. The protonation of the phospholipid may change at acidic pH: a pKa (COO–) of 5.5 and a pKa (PO2–) of 3–4 have been reported for PtdSer [37]. As APP undergoes a charge change from anionic to zwitterionic, only a part of PtdSer is protonated and still presents negative charges. Incorporation at pH lower than 4 was not analysed because these pH values do not correspond to physiological conditions. APP tends to adsorb on the air–water interface, and the increase in surface pressure was similar to that observed with other membrane or soluble proteins [38–40]. As reported by Graham and Phillips [34], the rate of the surface pressure change is a function of the stability of the protein structure, and more flexible molecules give a more rapid increase of the surface pressure. By analogy with other proteins [38,39], such a phenomenom suggests that APP may have a relatively flexible structure. CD experiments supported this hypothesis as the APP structure contains at least 30% of random coil in its native form [30]. Determination of the parameters that control the adsorption of APP at the air–water interface suggested that the molecular rearrangement at the interface of the monolayer bearing the absorbed protein depends on both the activation energy barrier of the insertion process and the bulk concentration of the protein. This observation is in agreement with protein adsorption models [38–40]. Under the experimental conditions used in this study, the interaction of APP with a neutral lipid monolayer is not dependent on the bulk pH, as it was observed for binding experiments into liposomes. These results may be explained by an effect of local interfacial pH [41]. Therefore, alteration of the charge of the protein could be induced at the vicinity of the monolayer. A difference in the orientation of the protein moiety during the insertion process into the monolayer film or into the bilayer containing the surfactant is also possible. The lower surface activity obtained with the PtdSer monolayers could be the result, as presented before, of an electrostatic repulsion effect between lipids and the protein at physiological pH. Anionic lipids, which represent
20% of biological membrane lipids, provide either a source of electrostatic


FEBS 2002

2244 R. Lahdo et al. (Eur. J. Biochem. 269)

attractions for the binding of proteins to membranes or a source of repulsive effects [42–44]. The results presented are consistent with the hypothesis that the binding of APP to phospholipid monolayers is also driven by hydrophobic interactions, as the increase of salt concentration favoured the penetration process of the protein into the monolayer. The binding of APP to phospholipid films could be linked to a decrease in solubility of the protein in the bulk phase at high ionic strength, leading to a greater amount of the protein present at the interface. In other words, increasing the ionic strength has the same effect as an increase of the protein hydrophobicity. As the hydrophobic nature of the interaction is stimulated by increasing the ionic strength of the subphase, the insertion rate of APP should be facilitated when the electrostatic repulsion between the lipid–protein molecules at the interface occurs. It may be the case for the interaction of APP with PtdCho monolayers. We investigated whether the observed changes are due to conformational changes of APP, associated with secondary, tertiary or quaternary modifications. Far-UV CD spectra of APP showed that the protein conformation is slightly affected by either the pH or the salt concentration, in the absence or in the presence of PtdCho vesicles. We concluded that the protein remains essentially in a a-helical structure under these conditions. We could not exclude, however, that local changes may occur and escape detection using CD, given the fact that APP is a large protein. These binding and CD studies show that the binding is not directly related to the secondary structure of the protein, as binding can be modified when secondary structure is not. However, it should be pointed out that CD spectra of the APP recorded at pH 4 suggested an aggregation and/or precipitation of the protein. The possibility that the lower surface activity of APP observed at pH 4 could be due to a lower available amount of the protein cannot be ruled out. No loss of tertiary structure, as measured by intrinsic fluorescence, indicates that the protein does not undergo major unfolding during acidification or upon variation of ionic strength (data not shown). Whether minor conformational modifications of APP, such as local variations in secondary or tertiary structures in defined regions of the protein, are associated with its lipid binding ability remains to be determined. Finally, we have considered the possibility that the pH-induced conformational transition may lead to changes in the protein quaternary structure. We have used gel filtration chromatography to test this hypothesis, and have obtained no evidence for a different structure at pH 4–7 (data not shown). The APP is apparently organized as a trimer in the conditions used for these studies. It was recently reported that cellular APP can form noncovalent homodimers and tetramers [45]. The molecular events which guide protein interaction at the membrane surface or insertion into lipid vesicles are important in order to shed light on the toxicity mechanism of Ab which may be based in part on perturbations of the lipid–water interface. In vivo, production of Ab is believed to occur through sequential cleavage of APP by b- and c-secretases [46]. Cell biological studies suggest three potential locations for intracellular b-secretase activity: endosomal compartments at mildly acidic pH, Golgiderived vesicles and endoplasmic reticulum/intermediate compartments. b-secretase has an acidic pH optimum of


4.5 [7,9,10]. The final step in the generation of Ab from the amyloid precursor protein is the proteolysis by the elusive c-secretase(s) [8,11,12]. From a biological point of view, the c-secretase cleavage is intriguing and unusual because it is an intramembranous cleavage. The predicted transmembrane domain of APP (from Gly625 to Leu648) ends just before the putative c-secretase cleavage site (Lys649–Lys651), which implies that the c-secretase cleavage sites are located in the hydrophobic environment of the cell membrane. The in vitro model where APP is incorporated into lipid layers could support the precision of the mechanism of proteolysis, and emphasize the lipid parameters that influence this process. At this point, it is interesting to note that numerous studies have detected extensive interactions between Ab and anionic lipids. More recently, different studies have suggested that cholesterol, an important determinant of the physical state of biological membranes, plays a significant role in the development of AD. Whether this particular lipid has an effect on the insertion or on the interaction of APP with lipid membranes is under investigation. Although we have shown that the interaction of the protein with neutral or anionic model membranes depends on electrostatic and hydrophobic effects, the participation of other factors should also be taken into consideration.

REFERENCES 1. Kang, J., Lemaire, H.-G., Unterbeck, A., Salbaum, J.M., Masters, C.L., Grzeschik, K.H., Multhaup, G., Beyreuther, K. & Mu¨llerHill, B. (1987) The precursor of Alzheimer’s disease amyloid A4 protein resembles a cell-surface receptor. Nature 325, 733–736. 2. Tanzi, R.E., Gusella, J.F., Watkins, P.C., Bruns, G.A., St GeorgeHyslop, P., Van Keuren, M.L., Patterson, D., Pagan, S., Kurnit, D.M. & Neve, R.L. (1987) Amyloid b protein gene: cDNA, mRNA distribution, and genetic linkage near the Alzheimer locus. Science 235, 880–884. 3. Selkoe, D.J., Podlisny, M.B., Joachim, C.L., Vickers, E.A., Lee, G., Fritz, L.C. & Oltersdorf, T. (1988) b-Amyloid precursor protein of Alzheimer disease occurs as 110- to 135-kilodalton membraneassociated proteins in neural and nonneural tissues. Proc. Natl Acad. Sci. USA 85, 7341–7345. 4. Neve, R.L., McPhie, D.L. & Chen, Y. (2000) Alzheimer’s disease: a dysfunction of the amyloid precursor protein. Brain Res. 886, 54–66. 5. Perez, R.G., Squazzo, S.L. & Koo, E.H. (1996) Enhanced release of amyloid b-protein from codon 670/671 Swedish mutant b-amyloid precursor protein occurs in both secretory and endocytic pathways. J. Biol. Chem. 271, 9100–9107. 6. Annaert, W.G., Levesque, L., Craessaerts, K., Dierinck, I., Snellings, G., Westaway, D., George-Hyslop, P.S., Cordell, B., Fraser, P. & De Strooper, B. (1999) Presenilin 1 controls c-secretase processing of amyloid precursor protein in pre-golgi compartments of hippocampal neurons. J. Cell. Biol. 147, 277–294. 7. Hussain, I., Powell, D., Howlett, D.R., Tew, D.G., Meek, T.D., Chapman, C., Gloger, I.S., Murphy, K.E., Southan, C.D., Ryan, D.M., Smith, T.S., Simmons, D.L., Walsh, F.S., Dingwall, C. & Christie, G. (1999) Identification of a novel aspartic protease (Asp 2) as b-secretase. Mol. Cell Neurosci. 14, 419–427. 8. Murphy, M.P., Hickman, L.J., Eckman, C.B., Uljon, S.N., Wang, R. & Golde, T.E. (1999) c-Secretase, evidence for multiple proteolytic activities and influence of membrane positioning of substrate on generation of amyloid b peptides of varying length. J. Biol. Chem. 274, 11914–11923.


FEBS 2002

Interaction of amyloid precursor protein with lipids (Eur. J. Biochem. 269) 2245

9. Sinha, S., Anderson, J.P., Barbour, R., Basi, G.S., Caccavello, R., Davis, D., Doan, M., Dovey, H.F., Frigon, N., Hong, J. et al. (1999) Purification and cloning of amyloid precursor protein b-secretase from human brain. Nature 402, 537–540. 10. Vassar, R., Bennett, B.D., Babu-Khan, S., Kahn, S., Mendiaz, E.A., Denis, P., Teplow, D.B., Ross, S., Amarante, P., Loeloff, R. et al. (1999) b-Secretase cleavage of Alzheimer’s amyloid precursor protein by the transmembrane aspartic protease BACE. Science 286, 735–741. 11. Esler, W.P., Kimberly, W.T., Ostaszewski, B.L., Diehl, T.S., Moore, C.L., Tsai, J.Y., Rahmati, T., Xia, W., Selkoe, D.J. & Wolfe, M.S. (2000) Transition-state analogue inhibitors of c-secretase bind directly to presenilin-1. Nature Cell Biol. 2, 428–434. 12. Li, Y.M., Xu, M., Lai, M.T., Huang, Q., Castro, J.L., DiMuzioMower, J., Harrison, T., Lellis, C., Nadin, A., Neduvelil, J.G., Register, R.B. et al. (2000) Photoactivated c-secretase inhibitors directed to the active site covalently label presenilin 1. Nature 405, 689–694. 13. De Strooper, B. & Annaert, W. (2000) Proteolytic processing and cell biological functions of the amyloid precursor protein. J. Cell Sci. 113, 1857–1870. 14. Simons, M., Keller, P., De Strooper, B., Beyreuther, K., Dotti, C.G. & Simons, K. (1998) Cholesterol depletion inhibits the generation of b-amyloid in hippocampal neurons. Proc. Natl Acad. Sci. USA 95, 6460–6464. 15. Galbete, J.L., Martin, T.R., Peressini, E., Modena, P., Bianchi, R. & Forloni, G. (2000) Cholesterol decreases secretion of the secreted form of amyloid precursor protein by interfering with glycosylation in the protein secretory pathway. Biochem. J. 348, 2307–2313. 16. Refolo, L.M., Malester, B., La Francois, J., Bryant-Thomas, T., Wang, R., Tint, G.S., Sambamurti, K., Duff, K. & Pappolla, M.A. (2000) Hypercholesterolemia accelerates the Alzheimer’s amyloid pathology in a transgenic mouse model. Neurobiol. Dis. 7, 321– 331. 17. Fassbender, K., Simons, M., Bergmann, C., Stroick, M., Lutjohann, D., Keller, P., Runz, H., Kuhl, S., Bertsch, T., von Bergmann, K., Hennerici, M., Beyreuther, K. & Hartmann, T. (2001) Simvastatin strongly reduces levels of Alzheimer’s disease b-amyloid peptides Abeta 42 and Abeta 40 in vitro and in vivo. Proc. Natl Acad. Sci. USA 98, 5856–5861. 18. Kojro, E., Gimpl, G., Lammich, S., Marz, W. & Fahrenholz, F. (2001) Low cholesterol stimulates the nonamyloidogenic pathway by its effect on the a-secretase ADAM 10. Proc. Natl Acad. Sci. USA 98, 5815–5820. 19. Mason, R.P., Shoemaker, W.J., Shajenko, L. & Herbette, L.G. (1993) X-ray diffraction analysis of brain lipid membrane structure in Alzheimer’s disease and b-amyloid peptide interactions. Ann. N. Y. Acad. Sci. 695, 54–58. 20. Simmons, L.K., May, P.C., Tomaselli, K.J., Rydel, R.E., Fuson, K.S., Brigham, E.F., Wright, S., Lieberburg, I., Becker, G.W., Brems, D.N. & Li, W.Y. (1994) Secondary structure of amyloid b-peptide correlates with neurotoxic activity in vitro. Mol. Pharmacol. 45, 373–379. 21. Buchet, R., Tavitian, E., Ristig, D., Swoboda, R., Stauss, U., Gremlich, H.U., de La Fournie`re, L., Staufenbiel, M., Frey, P. & Lowe, D. (1996) Conformations of synthetic b peptides in solid state and in aqueous solution: relation to toxicity in PC12 cells. Biochim. Biophys. Acta. 1315, 40–46. 22. Choo-Smith, L.-P., Garzon-Rodriguez, W., Glabe, C.G. & Surewicz, W.K. (1997) Acceleration of amyloid fibril formation by specific binding of Ab-(1–40) peptide to ganglioside-containing membrane vesicles. J. Biol. Chem. 272, 22987–22990. 23. Terzi, E., Holzemann, G. & Seelig, J. (1997) Interaction of alzheimer b-amyloid peptide (1–40) with lipid membranes. Biochemistry 36, 14845–14852.

24. Matsuzaki, K. & Horikiri, C. (1999) Interactions of amyloid b-peptide (1–40) with ganglioside-containing membranes. Biochemistry 38, 4137–4142. 25. Yip, C.M. & McLaurin, J. (2001) Amyloid-b peptide assembly: a critical step in fibrillogenesis and membrane disruption. Biophys. J. 80, 1359–1371. 26. Lin, H., Zhu, Y.J. & Lal, R. (1999) Amyloid b protein (1–40) forms calcium-permeable, Zn2+-sensitive channel in reconstituted lipid vesicles. Biochemistry 38, 11189–11196. 27. Mason, R.P., Jacob, R.F., Walter, M.F., Mason, P.E., Avdulov, N.A., Chochina, S.V., Igbavboa, U. & Wood, W.G. (1999) Distribution and fluidizing action of soluble and aggregated amyloid b-peptide in rat synaptic plasma membranes. J. Biol. Chem. 274, 18801–18807. 28. Kremer, J.J., Pallitto, M.M., Sklansky, D.J. & Murphy, R.M. (2000) Correlation of b-amyloid aggregate size and hydrophobicity with decreased bilayer fluidity of model membranes. Biochemistry 39, 10309–10318. 29. Vargas, J., Alarcon, J.M. & Rojas, E. (2000) Displacement currents associated with the insertion of Alzheimer disease amyloid b-peptide into planar bilayer membranes. Biophys. J. 79, 934–944. 30. de La Fournie`re-Bessueille, L., Grange, D. & Buchet, R. (1997) Purification and spectroscopic characterisation of b-amyloid precursor protein from porcine brains. Eur. J. Biochem. 250, 705–711. 31. Bradford, M.M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254. 32. Angrand, M., Briolay, A., Ronzon, F. & Roux, B. (1997) Detergent-mediated reconstitution of a glycosyl-phosphatidylinositol-protein into liposomes. Eur. J. Biochem. 250, 168–176. 33. Le´vy, D., Gulik, A., Bluzat, A. & Rigaud, J.L. (1992) Reconstitution of the sarcoplasmic reticulum Ca2+-ATPase: mechanisms of membrane protein insertion into liposomes during reconstitution procedures involving the use of detergents. Biochim. Biophys. Acta. 1107, 283–298. 34. Graham, D.G. & Phillips, M.C. (1979) Proteins at liquid interface. J. Colloid Interface Sci. 70, 403–414. 35. Pascher, I., Lundmark, M., Nyholm, P.G. & Sundell, S. (1992) Crystal structures of membrane lipids. Biochim. Biophys. Acta 1113, 339–373. 36. Parmar, M.M., Edwards, K. & Madden, T.D. (1999) Incorporation of bacterial membrane proteins into liposomes: factors influencing protein reconstitution. Biochim. Biophys. Acta 1421, 77–90. 37. van Dijck, P.W., de Kruijff, B., Verkleij, A.J., van Deenen, L.L. & de Gier, J. (1978) Comparative studies on the effects of pH and Ca2+ on bilayers of various negatively charged phospholipids and their mixtures with phosphatidylcholine. Biochim. Biophys. Acta 512, 84–96. 38. Maget-Dana, R. (1999) The monolayer technique: a potent tool for studying the interfacial properties of antimicrobial and membrane-lytic peptides and their interactions with lipid membranes. Biochim. Biophys. Acta 1462, 109–140. 39. Sun, Y., Wang, S. & Sui, S. (2000) Penetration of human apolipoprotein H into air/water interface with and without phospholipid monolayers. Colloids Surfaces A: Physicochem. Eng. Aspects. 175, 105–112. 40. de La Fournie`re, L., Ivanova, M.G., Blond, J.-P., Carrie`re, F. & Verger, R. (1994) Surface behavior of human pancreatic and gastric lipases. Colloids Surfaces B: Biointerfaces. 2, 585–593. 41. McLaughlin, S. (1989) The electrostatic properties of membranes. Annu. Rev. Biophys. Biophys. Chem. 18, 113–136. 42. Mel, S.F. & Stroud, R.M. (1993) Colicin Ia inserts into negatively charged membranes at low pH with a tertiary but little secondary structural change. Biochemistry 32, 2082–2089. 43. Nalefski, E.A., McDonagh, T., Somers, W., Seehra, J., Falke, J.J. & Clark, J.D. (1998) Independent folding and ligand specificity of

2246 R. Lahdo et al. (Eur. J. Biochem. 269) the C2 calcium-dependent lipid binding domain of cytosolic phospholipase A2. J. Biol. Chem. 273, 1365–1372. 44. Smith, E.R. & Storch, J. (1999) The adipocyte fatty acid-binding protein binds to membranes by electrostatic interactions. J. Biol. Chem. 274, 35325–35330. 45. Scheuermann, S., Hambsch, B., Hesse, L., Stumm, J., Schmidt, C., Beher, D., Bayer, T.A., Beyreuther, K. & Multhaup, G. (2001)


FEBS 2002 Homodimerization of amyloid precursor protein and its implication in the amyloidogenic pathway of Alzheimer’s disease. J. Biol. Chem. 276, 33923–33929. 46. Checler, F. (1995) Processing of the b-amyloid precursor protein and its regulation in Alzheimer’s disease. J. Neurochem. 65, 1431– 1444.