Insertion of Externally Administered Amyloid  Peptide 25-35 and Perturbation of Lipid Bilayers † Silvia Dante,* ,‡,§ Thomas Hauss, §,| and Norbert A. Dencher ‡ Physical Biochemistry, Darmstadt UniVersity of Technology, Petersenstrasse 22, D-64287 Darmstadt, Germany, Institute of Physical Biology, Heinrich-Heine-UniVersita ¨t, D-40225 Du ¨sseldorf, Germany, and BENSC, Hahn-Meitner-Institut, Glienicker Strasse 100, D-14109 Berlin, Germany ReceiVed June 18, 2003; ReVised Manuscript ReceiVed September 23, 2003 ABSTRACT: To understand the molecular basis and to prevent diseases such as Alzheimer’s disease (AD), the targets of the triggering agent have to be elucidated. -Amyloid peptide (A) is the major component of extracellular senile plaques characteristic of AD. For a very long time, the aggregated form of the A was supposed to be responsible for the neurodegeneration that occurs in AD. Recently, the attention has been diverted to the monomeric or oligomeric forms of A and their interaction with cellular targets. In our investigation, the physiological and medically important insertion of externally applied A monomers into the bilayer of lipid vesicles is demonstrated. A(25-35) has been localized in the region of the lipid alkyl chain, and it has a severe disordering effect on the lamellar order of the lipid bilayer. Both of these results are of biomedical relevance. Alzheimer’s disease (AD) 1 is characterized by the forma- tion of amyloid plaques, deposited in the neuronal extracel- lular space of the brain and surrounded by dystrophic neurons. The major component of these aggregates is A,a 39-42 amino acid peptide produced by enzymatic cleavage of a 770 amino acid membrane spanning precursor protein. The so-called amyloid hypothesis links the presence of the insoluble fibril-like A aggregates to the pathogenesis of the disease. For more than a decade the amyloid hypothesis has been the driving idea in the attempt to understand the trigger of the neuritic damage leading to dementia. A huge amount of data has been collected, in vivo and in vitro, that supports this theory; evidence in favor of it are for instance the death of neurons after administration of Α (1-3) and the observation of neurotoxicity of the fibrillar form of Α after intracerebral injection (4, 5). Other experimental evidence has been collected which does not support the assumption of neurotoxicity of fibrillar A. For instance, transgenic mice that accumulate a big load of plaques containing fibrillar A, show no or little neurotoxic response to this load (6). Other results were equivocal or not reproducible, this fact being ascribed to methodological differences (2, 3, 7, 8). Recently, new suppositions have been put forward. In a very provoca- tive article (9) Bishop and Robinson highlighted the weak- ness of the amyloid hypothesis and suggested that A might have a neuroprotective role instead (bioflocculant hypoth- esis). Other studies focus on nonfibrillar, shorter soluble forms of monomeric or dimeric A and show that they can be highly toxic. A strong correlation of the neurodegenerative process with the absolute level of soluble amyloid has been found (6). Moreover, soluble forms of A may be released from the senile mature plaques, interact with the neurons, and cause neuron damage or disruption (10). In this scenario, it is relevant to identify the target (e.g., a specific receptor or organelle or the lipid membrane itself) of A. It is also pertinent to investigate the interaction of monomeric A with lipid membranes, to clarify its possibility to intercalate in the lipid core, and to identify its location in the membrane. Very few direct methods have been applied for this purpose in the past. In this work, we have applied neutron diffraction in conjunction with selective amino acid deuteration to localize a short fragment of A, namely, A(25-35), in small unilamellar vesicles. A(25-35) (GSNKGAIIGLM) is considered to be the reactive part of A(1-42), since it shows fibrillogenic activity and has a neurotoxic effect on cultured cells. A(25-35) has channel-forming activity in bilayers (11). It was found to alter membrane fluidity in rat brain and in human cortex membranes (12, 13) and to modulate membrane lipid peroxidation (14). Its interaction with lipid layers has been studied with a variety of biochemi- cal techniques, and it was described to be dependent on the membrane composition, on the aggregation state of the peptide, and on pH and ionic strength (15-18). In previous papers the location of A(25-35) in oriented lipid bilayer samples was investigated by diffraction tech- niques, and the presence of the peptide in the hydrophobic core of the membrane was established (19-21). In the present study, for the first time we were able to prove by † This work was supported by Grant 03-DEE8DA from Bundesmin- isterium fu ¨r Bildung und Forschung, by the Fonds der Chemischen Industrie (to N.A.D.), and by the Research Center Ju ¨lich (F+E). * Corresponding author. E-mail: silvia.dante@hmi.de. Tel: +49 30 8062 2071. Fax: +49 30 8062 2999. ‡ Darmstadt University of Technology. § Hahn-Meitner-Institut. | Heinrich-Heine-Universita ¨t. 1 Abbreviations: A(25-35), -amyloid (25-35); AD, Alzheimer’s disease; POPC, 1-palmitoyl-2-oleoylphosphatidylcholine; POPS, 1-palm- itoyl-2-oleoylphosphatidylserine; TFA, trifluoroacetic acid; rh, relative humidity; H-Leu34-A(25-35), -amyloid (25-35) with protonated leucine in position 34; 2 H-Leu34-A(25-35), -amyloid (25-35) with deuterated leucine in position 34; fwhm, full width at half-maximum. 13667 Biochemistry 2003, 42, 13667-13672 10.1021/bi035056v CCC: $25.00 © 2003 American Chemical Society Published on Web 10/30/2003