Laser-Induced Liquid Bead Ion Desorption Mass Spectrometry: An

Article pubs.acs.org/ac

Laser-Induced Liquid Bead Ion Desorption Mass Spectrometry: An Approach to Precisely Monitor the Oligomerization of the β-Amyloid Peptide Mihaela Cernescu,† Tina Stark,‡ Elisabeth Kalden,‡ Christopher Kurz,§ Kristina Leuner,§ Thomas Deller,∥ Michael Göbel,‡ Gunter P. Eckert,§ and Bernhard Brutschy*,† †

Institute for Physical and Theoretical Chemistry, Goethe-University, Frankfurt/M, Germany, 60438 Institute for Organic Chemistry and Chemical Biology, Goethe-University, Frankfurt/M, Germany, 60438 § Department of Pharmacology, Biocentre, Goethe-University, Frankfurt/M, Germany, 60438 ∥ Institute of Clinical Neuroanatomy, Dr. Senckenberg Anatomy, Goethe-University, Neuroscience Center, Frankfurt/M, Germany, 60590 ‡

S Supporting Information *

ABSTRACT: In the present work, the recently developed laser-induced liquid bead ion desorption mass spectrometry (LILBID MS) is applied as a novel technique to study Aβ oligomerization, thought to be crucial in Alzheimer’s disease (AD). The characterization of the earliest nucleation events of this peptide necessitates the application of several techniques to bridge the gap between small oligomers and large fibrils. We precisely monitored in time the transformation of monomeric Aβ (1-42) into oligomeric Aβn (n < 20) and its dependence on concentration and agitation. The distribution shows signs of the hexamer being crucial in the assembly process. The intensity of the monomer decreases in time with a time constant of about 9 h. After a lag time of around 10 h, a phase transition occurred in which the total ion current of the oligomers goes to nearly zero. In this late stage of aggregation, protofibrils are formed and mass spectrometry is no longer sensitive. Here fluorescence correlation spectroscopy (FCS) and transmission electron microscopy (TEM) are complementary tools for detection and size characterization of these large species. We also utilized the oligomers of Aβ (1-42) as a model of the corresponding in vivo process to screen the efficacy and specificity of small molecule inhibitors of oligomerization. The LILBID results are in excellent agreement with condensed phase behavior determined in other studies. Our data identified LILBID MS as a powerful technique that will advance the understanding of peptide oligomerization in neurodegenerative diseases and represents a powerful tool for the identification of small oligomerization inhibitors.

A

their formation kinetics is one issue; another one is to inhibit Aβ aggregation by small molecule inhibitors. The latter may be a feasible therapeutic target for the treatment of AD. Thus, recent research has started to unravel oligomer structural and biophysical features.14 However, the precise details of the oligomerization process are still not fully understood.15 Aβ oligomerization is a concentration-dependent phenomenon5 in which monomeric Aβ initially forms poorly characterized oligomers. The formation of a paranucleus presumably is an important step in amyloid formation.16 Oligomers are known to induce the formation of protofibrils, which ultimately leads to the assembly of insoluble amyloid fibrils.4 In vitro studies have demonstrated that Aβ monomers can adopt the conformation of the α-helix, β-sheet, or random coil structure depending on the physical properties and chemical composition of the environment.17,18 Toxicity of Aβ is said to be related to β-sheet

lzheimer’s disease (AD) represents one of the most common late-onset, neurodegenerative disorders. Neuropathologically it is characterized by ordered protein aggregation of the Amyloid β-peptide (Aβ) forming neurofibrillary tangles and neuritic plaques. Aβ is composed of 38−43 amino acid residues, derived from the transmembrane amyloid precursor protein (APP) after sequential proteolytic cleavage by different secretases.1,2 The most common isoforms of the Aβ are the 40 and 42 amino acid peptides, Aβ (1-40) and Aβ (1-42), differing only in the truncation of two hydrophobic residues (I41 and A42) from the carboxylic terminus. Amphipathic Aβ in solution tends to self-aggregate and accumulate, which initiates a cascade that triggers complex pathological reactions eventually leading to neuronal dysfunction and cell death.1,3−5 Accordingly, current research suggests that soluble oligomeric forms of Aβ play a major role in AD pathophysiology.1,4,6,7 It has been shown that Aβ oligomers specifically interact with plasma membranes,8 interfere with the glutamatergic neurotransmission,9,10 disrupt dendritic spines,11 inhibit hippocampal longterm potentiation in vivo12 and provide neuron specific toxicity.13 To identify and characterize oligomeric species and © 2012 American Chemical Society

Received: January 26, 2012 Accepted: May 17, 2012 Published: May 17, 2012 5276

dx.doi.org/10.1021/ac300258m | Anal. Chem. 2012, 84, 5276−5284

Analytical Chemistry

Article

structures; however, there is not direct experimental evidence for this statement.19 It is not fully understood at which stage the first β sheets occur, and there is only a rough estimate about the size of the toxic species. Several studies have focused on the Aβ transition from a monomeric random-coil conformation to an aggregated β-sheet secondary structure. It has been shown that amyloid aggregation is a highly complex process depending on the primary sequence,20 the peptide concentration,21 the existence of seed fibrils,5 the ratio of monomeric to aggregated forms,22 the pH,23 membrane lipids,24 endogenous proteins,5 protein glycosylation,25 and transition metal ions.26 Different techniques to investigate Aβ oligomerization have been employed including native gel electrophoresis and silver staining,27 electron microscopy,28−30 fluorescence spectroscopy,31,32 fluorescence correlation spectroscopy,33 and mass spectrometry.16,34,35 MALDI mass spectrometric imaging (MSI)36 and detection of oligomers up to n = 8−10 together with a multimethodological approach37 are reported. In vitro, the use of MS is particularly useful when the substrate cleavage sites of different Aβ-degrading enzymes are to be identified.38,39 Electrospray ionization (ESI) combined with ion mobility (IM) was used to study Aβ(1-40) and Aβ(1-42) peptides in vitro, suggesting different oligomer distributions associated with the different fibrillation tendencies of the two Aβ forms.16,40 For this study, precise experimental protocols for producing and monitoring the protein complexes in solution are mandatory. The first challenge is to get reproducibility in the Aβ aggregation kinetics, the second challenge is the detection method, which should be able to detect the aggregates from their native environment, i.e., in vitro, the third challenge is very large oligomers or protofibrils, for which mass spectrometry is no longer appropriate. Here complementary methodological approaches have to be applied. In the present work, we introduce a novel approach to study Aβ oligomerization using the recently developed laser-induced liquid bead ion desorption (LILBID) MS technique.41 Ions are laser desorbed/ablated from micro droplets of aqueous solution containing the analyte and a specific buffer with a defined pH. Under carefully selected conditions, this method preserves specific protein−protein interactions so that both soluble and membrane protein complexes can be analyzed.42 Using LILBID, we precisely monitored the oligomerization process of Aβ and focused our study on the kinetics of the in vitro Aβ oligomer formation. At later times, oligomers are converted into larger species outside the range of MS where fluorescence correlation spectroscopy (FCS) and transmission electron microscopy (TEM) are used for characterization. We also monitored the formation of amyloid fibrils by a Thioflavine T assay. Moreover we also examined the efficacy of four small molecules to inhibit the process of oligomerization.

Figure 1. Preparation procedure: Lyophilized Aβ was dissolved in hexafluoroisopropanol (HFIP) α-helix stabilizer to avoid peptide aggregation. Aliquots were evaporated at room temperature and under vacuum using a SpeedVac. Dried Aβ peptide films were stored at −20 °C. For oligomer preparation, Aβ was dissolved in DMSO, diluted in a buffer of NH4CO3 (100 mM) at physiological pH, and stored at 4 °C until investigated by LILBID-MS, fluorescence correlation spectroscopy (FCS), and transmission electron microscopy (TEM) (see the Materials and Methods). These steps define the so-called standard conditions (STC).

Briefly, 1 mg of Aβ was dissolved in 200 μL of 1,1,1,3,3,3hexafluoro-2-propanol (HFIP) to 1 mM and then aliquoted into 10 μL portions in microcentrifuge tubes. The HFIP was allowed to evaporate in the fume hood, and the resulting clear peptide films were dried under vacuum for 45 min and stored desiccated at −20 °C. Preparation of Oligomeric Aβ Peptides. The dried films were resuspended in 2 μL of DMSO and diluted in 98 μL of 100 mM NH4CO3 buffer at pH 7.4 to achieve a working solution of 100 μM concentration. The solution was vortexed for 30 s and incubated at 4 °C for different times. Preparation of Fibrillar Aβ Peptides. These were prepared similar as the oligomeric peptides but at 37 °C. We have recently characterized these Aβ preparations using native gel electrophoresis, followed by silver staining and electron microscopy.43 Cy5-Labeling of Aβ. The procedure of the Cy5-labeling is described in the Supporting Information in the Materials and Methods section. Laser-Induced Liquid Bead Ion Desorption Mass Spectrometry. The formation of early state Aβ oligomers was studied using laser-induced liquid bead ion desorption mass spectrometry (LILBID-MS), considered as an amalgamation of MALDI- and ESI-MS.44 The method uses liquid droplets of solution,42 which minimizes the analyte consumption for a measurement to typically a few microliters of solution. The micro droplets (diameter = 50 μm) are produced on demand by a piezo-driven droplet generator and introduced by differential pumping stages into the vacuum, where they are irradiated one by one by infrared laser pulses.41 The ions ejected are analyzed by time-of-flight MS. Thioflavin T Assay. The fluorescent cationic benzothiazole dye, ThT, is known to show increased fluorescence upon

■

MATERIALS AND METHODS Unless otherwise stated, reagents including 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP) were from Sigma-Aldrich (Taufkirchen, Germany). Human Aβ1‑42 (catalog number H-1368) was obtained from Bachem (Heidelberg, Germany). Preparation of Aβ1‑42 Peptides. We restricted our study mainly to Aβ1‑42. The preparation of oligomeric and fibrillar Aβ1‑42 was performed as previously reported,27,43 termed “preparation under standard conditions (STC)” (Figure 1). Since we are studying mainly Aβ1‑42, it is abbreviated as Aβ in the following. 5277



Article

Figure 2. Tracking of Aβ oligomerization using LILBID-MS. (A) Anion mass spectra of the Aβ oligomers after different incubation times. The Aβ peptide solutions prepared at STC were stored at 4 °C and gently mixed before sampling at the indicated time points (0, 4, 8, 24 h). For every spectrum, signals from 200 droplets were averaged. The fit of the peak maxima is indicated by the dashed line. Multiples of the singly charged pentamers are indicated. (B) Normalized monomer depletion diagram normalized to a concentration standard. The intensity was calculated by the peak integral from the LILBID-MS spectra. Experiments were repeated for three different samples. Mean values ± SEM were fitted by an exponential function with a time decay constant of around 9 h.

type Aβ peptide (DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVIA-42) carries a theoretical net charge of −3 and is relatively soluble.34 Hence for the present study, the MS measurements were done in the anion mode with only anions being analyzed. The oligomerization in solution was monitored at different times by sampling very small amounts of Aβ solution prepared under standard conditions (STC) and by mass analyzing the Aβ oligomer distribution by means of LILBID-MS. The standard protocol used for the preparation of Aβ solutions is depicted in Figure 1. The measurements under relatively quiescent conditions were done at 0, 4, 8, and 24 h after the initial incubation of the sample in a defined environment, as depicted in Figure 2A. The signal-to-noise ratio of the spectra and the spectrometer sensitivity allowed the identification of different oligomers ranging from the monomer up to the 20-mer (i.e., from ∼4 to 90 kDa). Apart from the dominant ion distribution, which corresponded to singly charged Aβ oligomers (Aβ)n‑, a distribution with much lower intensity was assigned to doubly charged oligomers appearing between the singly charged species. Mass spectra were recorded in analog mode using a fast transient recorder for digitizing the TOF-spectra. The optimal laser intensity for observing the oligomer distribution corresponded to a medium energy of 5 mJ/pulse for the desorption laser where larger oligomers could be distinguished. For the laser dependence of the ion signal, see the data in the Supporting Information (Figure S-1). In order to improve the signal-to-noise ratio, the mass spectra were averaged over 200 droplets. In addition, we used bovine serum albumin (BSA), as a concentration standard (585 amino acids, molecular mass 66 kDa) and we normalized the ion signals to the intensity of the triply charged BSA peak (5 μM). This is motivated by the fact that the intensity of the desorption laser may vary and with it the absolute detection efficiency. In LILBID at a constant laser intensity, the ion signals are strictly proportional to the concentration of the analyte in a range from 100 nM to 100 μM.42 An unstructured background in the ion signal may happen due to an incomplete or

binding to amyloid fibrils. When it binds to β sheet-rich structures, such as those in amyloid aggregates or fibrils, the dye displays enhanced fluorescence and a characteristic red shift of its emission spectrum. Hence the ThT assay is commonly used to quantify the formation of amyloid fibrils, both ex vivo and in vitro. Fluorescence Correlation Spectroscopy. To detect the conversion of Aβ oligomers into early aggregates and to observe their further enlargement, standard fluorescence correlation spectroscopy (FCS) was applied. Transmission Electron Microscopy. Transmission electron microscopy (TEM) was used for the visualization of large Aβ peptides aggregates as previously described.43 For more details about the above listed methods see the Supporting Information Materials and Methods section. Synthesis of Peptide Inhibitors. The inhibitors OR-1, OR-2, and the C-terminal peptide were prepared by conventional solid phase peptide synthesis using Fmoc amino acids. The dipeptide (D)Trp-Aib was synthesized by direct coupling of Fmoc-(D)Trp-OH with Aib-OMe in solution followed by removal of the protecting groups (see the Materials and Methods section in the Supporting Information). Statistics. For consistency, all experiments were repeated at least three times. Statistical analyses were performed using oneway ANOVA followed by a Tukey comparison test. The correlations were calculated using the Pearson GraphPad Prism 5.0 software package (San Diego, CA). In the LILBID-MS spectra, the Aβ oligomer populations were determined by performing peak integration using the Matlab 7.0 software package and their statistical approaches were done using Origin software.

■

RESULTS Tracking of Aβ Oligomerization Using LILBID-MS. Aggregation of Aβ strongly depends on the environment. One of the key factors in this context is the pH value, determining the peptide charge. For our measurements, at pH 7.4, the wild5278



Article

metastable desolvation. It was subtracted to determine the genuine signal peaks. The spectra were processed and plotted using Origin software. The relative population of each n-mer was determined by the integral over the corresponding signal. Unfortunately, the signal-to-noise ratio (S/N) drops for higher order oligomers. From the recorded spectra at 100 μM (Figure 2A), it is evident that the Aβ oligomers are formed shortly after dilution of the protein into the assembly buffer (0 h), corresponding to typically a few minutes. The shift in the peak intensities over time toward higher masses indicates that the Aβ oligomers do not grow by fusion but rather by attachment of monomers. Figure 2B depicts the BSA normalized signal of the monomer over time. These values were fitted by a monoexponential decay function, giving a time constant of τ = 9 ± 3 h. Rather different aggregation kinetics were observed when the stock solution was vortexed before each sampling. In Figure 3, larger oligomers (up to 20-mer) can be observed in the MS spectrum up to 8 h after sample preparation. At later times (10−24 h) these larger oligomers were found to be depleted.

Figure 4. Normalized total current of the n-mers (n > 1) (■) for different incubation times (mean value ± SEM) and normalized ThT fluorescence (□). The rapid decay of the current above 8−10 h is fitted by an exponential with a decay constant of 3.5 h.

for the formation of larger molecular species. Since these species do not appear in the MS, we assumed that their concentration is too low and/or their mass too large for their detection with LILBID. Hence we expected species of considerable size. In order to determine the critical concentration of Aβ where the aggregation starts, we also measured the dependence of the observed oligomers as a function of the Aβ concentration, ranging from 2 to 100 μM. After normalization to BSA, the Aβ monomer intensity and the total current of the oligomers were plotted against the concentration of Aβ (Figure S-3 in the Supporting Information). From these experiments, we deduce that the critical concentration where aggregation starts is below 5 μM. Tracking of Aβ Oligomerization Using FCS. FCS (fluorescence correlation spectroscopy) is a single molecule tracking technique to determine the dynamic properties of dyelabeled molecules in solution. It measures fluctuations of the fluorescence F(t) while the labeled molecule is diffusing through the focal volume of the excitation laser (∼1 fL). To detect only a few fluorescent molecules, their concentration must be in the nanomolar range.45 From the calculated autocorrelation function of F(t), the diffusion time of the observed species is determined. The software of the commercial instrument can monitor up to three different species independently by analysis of the autocorrelation function provided; they differ in molecular mass by at least 1 order of magnitude. Since the mobility depends on the mass and shape of the diffusing molecule, it is influenced by complexation. Hence if a fluorescent ligand binds to a macromolecule, this interaction can be followed by the change in diffusion time.46 Oligomerization and aggregation can be observed using FCS by an increase of the diffusion constant and inhibition by a decrease or constancy.47,48 Samples of Aβ (100 μM) containing 0.5 μM of Cy5-labeled Aβ were analyzed by FCS after 10-fold dilution. In contrast to LILBID, this method cannot resolve the exact oligomer composition. At t0, the best fit to the autocorrelation function was obtained by assuming the presence of a minor second “species”. Since a broad distribution in particle size occurs upon aggregation of Aβ, the dominant light component (diffusion time τ = 80 μs) should be seen as a mixture of monomer and

Figure 3. Tracking of Aβ oligomerization using LILBID-MS. Anion mass spectra of the Aβ oligomers from 0 to 24 h. The peptide solution was vortexed before each sampling at the indicated time points (vortexed mode). For every MS spectrum, signals from 200 droplets were averaged. At longer incubation times (>10 h), the higher oligomers decrease in intensity and disappear at 24 h. This may be explained by the rapid formation of protofibrils and fibrils with a concentration too low to be detected.

The total current (sum over the whole distribution) for the sample vortexed before each analysis was roughly constant up to around 10 h and then follows an exponential decay curve (Figure 4) (τ = 3.5 h). As the region where the total current is decreasing corresponds to an increase in the ThT fluorescence signal (see Figure 4), we may assume that this is an indication 5279



Article

100 000 μs, corresponding to 500−1500mers in the size range of micrometers). As depicted in Figure 5, the small Aβ oligomers decreased in intensity in time, while medium size and large species with longer diffusion times were progressively formed. The depletion of the light fractions follows a similar time course as that of the monomer in LILBID (Figure 2B). Independent from the development of diffusion times (Table S1 in the Supporting Information), a steady decrease in the detected number of particles was also found, indicating the formation of large species. Characterization of Aβ Peptides Using TEM. The gradual disappearance of larger Aβ oligomers in a vortexed sample (Figure 3) was rationalized by high molecular weight particles with large mass and very low concentration, which cannot be detected with LILBID. To provide evidence for this hypothesis, we followed the formation of larger particles from oligomeric to fibrillar Aβ formation using electron microscopy. Figure 6 shows the particle distributions of samples analyzed at different times after incubation. Increasingly larger particles were formed over time (Figure 6A,B). It should be noted that in all preparations a small number of large particles were observed shortly after dissolving the sample (0 h) which may possibly act as aggregation nuclei. Small Molecules Inhibit the Formation of Aβ Oligomers. The ability of LILBID-MS to detect low mass oligomers motivates its potential use for in vitro screening of compounds that inhibit oligomerization. Hence we tested four small molecules, known from the literature to block Aβ

smaller oligomers whereas the second “species” (best fit for τ around 5000 μs) represents larger oligomers consisting of 50− 100 subunits with lengths in the range of 0.1 μm (Figure 5; for

Figure 5. Characterization of Aβ oligomers during the assembly time course using FCS. Size distribution of oligomeric Aβ species as a function of time. The light gray column (diffusion time 1 in Table S-2 in the Supporting Information) corresponds to monomers and small oligomers, the dark gray column (diffusion time 2) to medium size oligomers (protofibrils), and black stands for large aggregates (fibrils) (diffusion time 3).

size calculations see the Supporting Information). After incubation times beyond 4 h, the best fit to the autocorrelation function was obtained by assuming a third species (τ = 40 000−

Figure 6. Formation of medium and large sized particles (protofibrils and fibrils) after different incubation times (transmission electron microscopy). (A) TEM-micrographs of Aβ oligomer (4 °C) and fibrillar (37 °C) preparations (magnification, 3.000; scale bar, 5 μm) illustrating the formation of protofibrils (2−5 μm) and fibrils (>5 μm) within 24 h. (B) TEM-micrographs of Aβ oligomer- and fibrillar preparations (magnification, 20.000; scale bar, 0.75 μm) for the time points 0 and 24 h, respectively. (C) Number of protofibrils and fibrils counted on electron micrographs (magnification, 3000) of Aβ oligomer preparations (mean ± SD; n = 3). 5280



Article

OR2. OR2 with the same amino acid sequence as OR1 carries a carboxamide at the C-terminus.50 LILBID-MS spectra of mixtures of Aβ with OR2 (1:4) at different time intervals after sample preparation show apart from monomers and dimers, only adducts with OR2 even after 24 h (Figure 7A). The spectra look unchanged after 48 and 72 h. This indicates a permanent inhibition of the oligomer formation. These findings confirm results from the condensed phase.50 C-terminal fragments (CTFs) have also been shown to strongly inhibit Aβ induced neurotoxicity.40,49 We studied a special fragment (Aβ (39-42), VVIA), which was synthesized together with three Arg residues to increase solubility (CTF, VVIARRR). The resulting oligomer distribution is depicted in Figure 7B. Apart from Aβ and its adducts with CTF, no larger oligomers appear. No change in the oligomeric state was observed the following days. (D)Trp-Aib. Another type of oligomerization inhibitors are so-called β-sheet breaker peptides which are similar to Aβ in sequence and hydrophobicity but contain amino acids like proline and 2-methylalanine (aminoisobutyric acid, Aib). It has been reported in the literature that this inhibitor causes a low propensity to form β-sheets in oligomerization.45,51 This peptide had the sequence (D)Trp-Aib.52 Unfortunately, from the time dependent LILBID-MS spectra, it is obvious that (D)Trp-Aib is not able to inhibit Aβ oligomerization and after 24 h a broad distribution of higher oligomers is visible in the MS spectra (Figure 7A). The inhibitors OR2, (D)Trp-Aib, and the C-terminal peptide were further investigated with FCS to confirm the results of LILBID in the condensed phase (Figure S-4 in the Supporting Information). This method, although unable to discriminate between different oligomers, confirm that OR2 and CTF are highly efficient blockers of aggregation, whereas (D)Trp-Aib showed no effect, in full accord with the findings by LILBIDMS.

aggregation.45−50 The samples were prepared at STC (Figure 1) with the only change being the aggregation buffer now containing the inhibitor (molar ratio of 1:4). To optimize the observation of oligomers, the samples were gently handled (quiescent mode). OR1. The first inhibitor peptide tested was OR1,50 (m = 1077 Da). It contains the amino acid sequence of Aβ responsible for the self-association (KLVFF, residues 16−20). In addition, RG and GR residues were added at the N- and Cterminal ends, respectively, which confer two positive net charges to OR1 to increase the peptidés solubility. Time dependent LILBID-MS spectra of the Aβ−OR1 sample were recorded at 0 h and 24 h as shown in Figure 7B. At 0 h,

■

DISCUSSION Detection of larger Aβ oligomers by mass spectrometry remains a technical challenge. With ESI only smaller oligomers are resolved,16 with MALDI species up to the decamer were detected, but here the environment is not native. The sensitivity of LILBID-MS allowed detection of (Aβ)n‑ oligomers (n < 20) and to follow the time dependent aggregation. For the statistical analysis of the presented spectra, all experiments were repeated at least three times. By using LILBID-MS, we could identify early Aβ oligomers, which are forming shortly after sample preparation. Moreover, we have observed two different aggregation kinetics depending on sample manipulation. When the sample was gently handled (quiescent mode) (Figure 2A), we could observe how peak intensities shift toward higher masses indicating that the low size Aβ oligomers continuously grow in size. A different aggregation behavior was observed when the solution was vortexed before each sampling (vortexed mode) (Figure 3). Then the oligomerization process was accelerated, and 24 h after sample preparation, only low mass species (n < 4) of decreasing intensity characterize the spectrum. This can be rationalized by assuming that stirring the solution enhances the association rates of Aβ peptides. It is clear that by a fast nucleation of larger oligomers to protofibrils or fibrils the concentration of the latter must be much lower than of the small oligomers.

Figure 7. Test of inhibitor peptides. (A) Normalized LILBID-MS spectra of the Aβ in the presence of the OR1 peptide in a 1:4 molar ratio. After 24 h incubation at 4 °C, the overall signal intensity is decreasing indicating the formation of oligomers (see magnification). (B) Normalized LILBID-MS spectra of Aβ in ascending order in the presence of the (D)Trp-Aib, OR2 ,and CTF, mixed in a 1:4 molar ratio and sampled after 24 h at 4 °C. The mixed complex formation is highlighted by arrows (1:1, 1:2, and 1:3). For every MS spectrum, signals from 200 droplets were averaged.

apart from the singly charged Aβ monomer and dimer peaks, various complexes between Aβ and OR1 peptides could be observed (1:1, 1:2, 1:3, 2:1, 2:2). At 24 h, the distribution shows small sized adducts of Aβ oligomers with OR1. Larger complexes were negligible. After 48 and 72 h, no change occurred in the oligomer distribution. This observation confirms the previous findings50 that the OR1 peptide inhibits the fibril but not the oligomer formation. 5281



Article

In LILBID the laser intensity is the parameter controlling fragmentation of weakly bound aggregates; however, we did not see the appearance of magic oligomers by reducing the laser intensity down into the threshold region (Figure S-1 in the Supporting Information). In an ESI setup as utilized by Bernstein et al. inelastic collisions happen in the ion-funnel, where ions collide with residual gas. For example, because of these collisions the ions observed in the ESI-MS setup are in general completely dehydrated. Hence by collision induced excitation of the oligomers these may partially be annealed, to give the most stable conformers.16 This may perhaps explain why only the hexamer and its fragmentation products tetramer, dimer, or its association product dodecamer are detected in ESI. In any case, the hexamer and its multimers seem to be particularly stable species. Does that give evidence that in liquid also only particles from this cascade exist? Hexamers were also observed in the in vitro snapshots by PICUP but pentamers and heptamers of similar intensity appeared as well. So these different oligomers seem to be also formed in solution. In LILBID, ions other than in ESI are formed in the laser induced explosive expansion of the droplet, whereby preformed solvated ions escape into vacuum by incomplete ion-recombination (lucky survivor model). The ions are in general not dehydrated and are stabilized by evaporative cooling. It seems that both PICUP and LILBID show a structurally metastable but important region around the hexamer. In LILBID after 24 h, when a substantial amount of Aβ has disappeared by aggregation (see the discussion below), the pentamer/hexamer is the dominant species as one expects for a kinetic bottleneck. In the vortexed mode this maximum does no longer build up probably by a fast depletion of the hexamer by accelerated formation of prefibrilar assemblies. So from the LILBID results, there is also no doubt that the hexamer plays an important role in the oligomerization process, as already identified by PICUP and ESI-MS. Since we have no ion-mobility spectrometer we assume that the oligomers observed are mainly singly charged, since doubly charged species with uneven aggregation number appear only at very low intensity (1−2%) between the singly charged main peaks. We also measured for control the size distribution of the oxidized alloform Met35 (O), since it was reported by the above authors that the oxidized polar side chain of Met(O) or Met(O2) disfavors the formation of a paranucleus.56 Our spectra showed besides the monomer, dimers and tiny amounts of trimers and tetramers over the whole range of incubation times (Figure S-7 in the Supporting Information). Thus similar to the PICUP and ESI studies, we found no evidence for an efficient oligomerization and hence a very different behavior than that of WT Aβ42. It was reported that some amyloid fibrils assemble via nucleated growth, involving the formation of a critical, highenergy nucleus during a lag phase, after which protofibril formation occurs very fast. The observation that vortexing increases the association rate of larger assemblies could be rationalized that it breaks larger oligomers to form larger numbers of nuclei to which the peptides will associate with the same rate. In our measurements, such a phase transition can be deduced from the rapid decrease in the total current over time. The onset time of this decrease would correspond to a lag time. In order to prove this hypothesis, we examined the formation of larger particles into multioligomeric or fibrillar species using FCS, TEM, and a ThT assay and the results are consistent with the conclusions from the LILBID spectra.

The oligomer distribution needs detailed discussion. The distributions in Figures 2A and 3 show no magic numbers, i.e., no discrete, dominant peaks appear. On the other side some peculiarities show up under closer inspection. After 24 h in quiescent mode, there is a maximum between the tetramer and hexamer, while for the shorter times the distribution shows rather evenly distributed species of decreasing intensity with size. When one calculates the second derivative of these distributions (Figure S-5 in the Supporting Information), with progressing time there starts to develop an inflection point at the hexamer, corresponding to the maximum at 24 h. Distributions sampled in the vortexed mode (Figure 3) on the other side no longer exhibit a maximum but at 8 h and 10 h only a shoulder in the region of the former maximum extending from dimer to decamer while for other times this change in intensity is not that prominent. The second derivative (Figure S-6 in the Supporting Information) after falling steeply from monomer to tetramer increasingly levels-off for larger oligomers with time. So in both modes there is a marked change in the aggregation behavior from tetramer to hexamer which is more pronounced in the quiescent mode. In the literature, mainly two groups report discrete oligomer sizes. Bitan et al.53 utilized among different methods for determining in vitro oligomer size distributions, a method called photoinduced cross-linking of unmodified proteins (PICUP). By this cross-linking-technique it was possible to take snapshots of the oligomer size distribution immediately after preparation of the amyloid. Afterward the stable, cross-linked species were analyzed by SDS-PAGE. The results revealed three groups of oligomers: a first one decreasing from monomer to trimer, a second group from tetramer to heptamer showing higher intensities and a maximum at the pentamer/hexamer, and a third less intensive group with oligomers of mass Mr ≈ 30 000− 60 000 with small maxima for nonamer and dodecamer. Larger species have also been found but only by dynamic light scattering.53 The region pentamer/hexamer was assigned to formation of a paranucleus that form the basic unit that associates further to form larger assemblies. A morphological analysis by electron microscopy without cross-linking revealed small 5 nm structures, probably spheroidal, which appear either individually or associated into small groups. After cross-linking, the structures are connected to each other by narrow threads. Similar discrete structures were observed by Bernstein et al. by electrospray (ESI) mass spectrometry coupled to an ionmobility spectrometer.34 In the negative ESI spectrum, only peaks at m/z = −4, −3, −5/2, and −2 appeared. The peak at m/z = −3 was identified as a (monomer)−3. Surprisingly ionmobility measurements revealed that the peak at m/z = −5/2 was both from a (dimer)5−, a (tetramer)10−, a (hexamer)15−, and a (dodecamer)30− of Aβ42. Hence, different oligomers with charge states giving the same m/z contribute to the signal. In particular, no pentamers or octamers were found. Larger oligomers have also been found later with a high/mass quadrupole time-of flight (QTOF) but could not be assigned to certain sizes. Different from these data determined with PICUP and ESI, LILBID spectra show no magic sizes but a change in the intensity distribution in the region of the hexamer similar to the findings of the above methods. Possible reasons must be considered. Both LILBID and ESI are soft methods.54,55 Both are on one side able to analyze noncovalently bonded complexes, but on the other side often induce a certain amount of fragmentation in case of very weakly bound species. 5282



Article

The action of the C-terminal peptide was tested at 24 h (Figure 7B). Apart from Aβ monomers and Aβ−CTF complexes, no large oligomers could be observed. On the other side, the so-called β-sheet breaker peptide45,51 (D)TrpAib52 unexpectedly showed no inhibition of Aβ oligomerization even after 24 h. Moreover, no mixed complexes between Aβ and (D)Trp-Aib could be observed (Figure 7B). These findings were also confirmed by the FCS measurement. In conclusion, our data identified LILBID-MS as a powerful technique that will advance the understanding of peptide oligomerization in neurodegenerative diseases and represents a powerful tool for the identification of small oligomerization inhibitors.

From the dependence of the total current as function of the concentration of Aβ (Figure S-3A in the Supporting Information), it should be pointed out that below 20 μM immediately after sample preparation (0 h), no oligomers can be identified. At 24 h later, oligomers become visible for concentrations larger than 10 μM, suggesting that the aggregation is taking place also for one-digit micromolar concentrations of Aβ. The association rate increases with concentration. Moreover, after 24 h the relative percentage of larger species, corresponding to a depletion of the total current, increases also with concentration. The formation of oligomers in the concentration range studied is also revealed by reduced monomer intensity after 24 h (Figure S-3B in the Supporting Information). We deduce for the critical concentration where aggregation starts, a value less than 5 μM. Furthermore, we tested four small molecules, which are already known from the literature to block Aβ aggregation. Two peptides, designed from the binding region of Aβ (KLVFF, residues 16-20) were tested for their effect on oligomerization and toxicity. 5 0 The OR2 peptide (RGKLVFFGR-NH2) proved to have the best inhibitory effect when compared with OR1 peptide (RGKLVFFGR), which showed only inhibition of larger aggregates. CTF peptides from the C-terminal end of Aβ are also discussed as a possible inhibitor of the aggregation in the literature.49 A typical compound with the last four amino acids of Aβ, VVIA, coupled with three arginines to increase solubility was also successfully tested for its inhibition potential. In LILBID-MS, monomers and dimers of Aβ as well as their adducts with OR2 peptides are dominating the recorded mass spectra and no change in this oligomer distribution was observed several days after incubation (Figure 7B). These results were also supported by the FCS measurements. For OR1, however, after 24 h a distribution of small sized Aβ oligomers (up to decamer) are detected, containing OR1 (Figure 7A). Thus our results confirm the previous findings50 that the OR1 peptide inhibits the formation of large species but not the formation of smaller oligomer adducts. In discussing ion intensities, one has to be aware that in LILBID individual molecules may have a different detection probability in the liquid to vacuum transition (transfer function). We assume for the ionization mechanism an incomplete neutralization of the preformed ions which depends on their charge state in solution.41 Although there is only a propensity for higher detection efficiency with higher charge state of the ions in solution, this effect should be considered in interpreting the mass spectra. Since an Aβ monomer has a net charge of −3 and the tested inhibitor peptides have 3 (in the case of OR1 peptide) or 4 positive charges (in the case of OR2 and CTF, respectively), the theoretical net charge of the mixed complexes becomes 0 or +1. So one would expect a significantly reduced detection probability for (Aβ−inhibitor) complexes as compared to Aβ monomer. This would explain in the case of inhibitors why adducts do not dominate the LILBID spectrum. This effect of the net charge was also tested with triple arginine as a noninhibiting ligand, where the spectrum was only reduced in intensity by reduction of the charge state but the oligomer distribution was more or less unchanged apart from additional adduct peaks (Figure S-2 in the Supporting Information). For the same reason we believe that the relative concentration in solution of the inhibitor complexes are substantially larger than their relative peak intensities in the recorded LILBID-MS spectra.

■

ASSOCIATED CONTENT

S Supporting Information *

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS M.C. and T.S. contributed equally to this work. We gratefully acknowledge stimulating discussions with Prof. Jean-Pierre Schermann and the technical assistance by Dr. Hans-Dieter Barth. We also acknowledge the help of Dr. Ekaterini Copanaki and Anke Biczysko. This work was supported in part by grants from the Alzheimer Forschung Initiative e.V. (AFI No. 08823 to G.P.E.) and the Frankfurt Cluster of Excellence, “Macromolecular Complexes” (B.B. and M.C.).

■

REFERENCES

(1) Haass, C.; Selkoe, D. J. Nat. Rev. Mol. Cell. Biol. 2007, 8, 101− 112. (2) Selkoe, D. J. Neuron 1991, 6, 487−498. (3) Hardy, J.; Selkoe, D. J. Science 2002, 297, 353−356. (4) Golde, T. E.; Dickson, D.; Hutton, M. Curr. Alzheimer Res. 2006, 3, 421−430. (5) Harper, J. D.; Lansbury, P. T., Jr. Annu. Rev. Biochem. 1997, 66, 385−407. (6) Shankar, G. M.; Li, S.; Mehta, T. H.; Garcia-Munoz, A.; Shepardson, N. E.; Smith, I.; Brett, F. M.; Farrell, M. A.; Rowan, M. J.; Lemere, C. A.; Regan, C. M.; Walsh, D. M.; Sabatini, B. L.; Selkoe, D. J. Nat. Med. 2008, 14, 837−842. (7) Shankar, G. M.; Bloodgood, B. L.; Townsend, M.; Walsh, D. M.; Selkoe, D. J.; Sabatini, B. L. J. Neurosci. 2007, 27, 2866−2875. (8) Eckert, G. P.; Wood, W. G.; Muller, W. E. Curr. Protein Pept. Sci. 2010, 11, 319−325. (9) Renner, M.; Lacor, P. N.; Velasco, P. T.; Xu, J.; Contractor, A.; Klein, W. L.; Triller, A. Neuron 2010, 66, 739−754. (10) Ronicke, R.; Mikhaylova, M.; Ronicke, S.; Meinhardt, J.; Schroder, U. H.; Fandrich, M.; Reiser, G.; Kreutz, M. R.; Reymann, K. G. Neurobiol. Aging 2011, 32, 2219−2228. (11) Lacor, P. N.; Buniel, M. C.; Furlow, P. W.; Clemente, A. S.; Velasco, P. T.; Wood, M.; Viola, K. L.; Klein, W. L. J. Neurosci. 2007, 27, 796−807. (12) Walsh, D. M.; Klyubin, I.; Fadeeva, J. V.; Cullen, W. K.; Anwyl, R.; Wolfe, M. S.; Rowan, M. J.; Selkoe, D. J. Nature 2002, 416, 535− 539. (13) Ebenezer, P. J.; Weidner, A. M.; Levine, H., III; Markesbery, W. R.; Murphy, M. P.; Zhang, L.; Dasuri, K.; Fernandez-Kim, S. O.; 5283



Article

Bruce-Keller, A. J.; Gavilan, E.; Keller, J. N. J. Alzheimers Dis. 2010, 22, 839−848. (14) Stefani, M. FEBS J. 2010, 277, 4602−4613. (15) Ghosh, P.; Kumar, A.; Datta, B.; Rangachari, V. BMC Bioinf. 2010, 11 (Suppl 6), S24. (16) Bernstein, S. L.; Dupuis, N. F.; Lazo, N. D.; Wyttenbach, T.; Condron, M. M.; Bitan, G.; Teplow, D. B.; Shea, J. E.; Ruotolo, B. T.; Robinson, C. V.; Bowers, M. T. Nat. Chem. 2009, 1, 326−331. (17) Liu, D.; Xu, Y.; Feng, Y.; Liu, H.; Shen, X.; Chen, K.; Ma, J.; Jiang, H. Biochemistry 2006, 45, 10963−10972. (18) Mandal, P. K.; Pettegrew, J. W. Neurochem. Res. 2004, 29, 2267−2272. (19) Masafumi, S.; Tamotsu, Z. FEBS J. 2010, 277, 1348−1358. (20) Fraser, P. E.; Nguyen, J. T.; Inouye, H.; Surewicz, W. K.; Selkoe, D. J.; Podlisny, M. B.; Kirschnr, D. A. Biochemisty 1992, 31, 10716− 10723. (21) Burdick, D.; Soreghan, B.; Kwon, M.; Kosmoski, J.; Knamer, M.; Henshen, A.; Yates, J.; Cotamn, C.; Glabe, C. J. Biol. Chem. 1992, 267, 546−554. (22) Jan, A; Gokce, O.; Luthi-Carter, R.; Lashuel, H. J Biol. Chem. 2008, 283, 28176−28189. (23) Fraser, P. E.; Nguyen, J. T.; Surewicz, W. K.; Kirshner, D. A. Biophys. J. 1991, 60, 1190−1201. (24) Ambroggio, E. E.; Kim, D. H.; Separovic, F.; Barrow, C. J.; Barnham, K. J.; Bagatolli, L. A.; Fidelio, G. D. Biophys. J. 2005, 88, 2706−2713. (25) Vitek, M. P.; Bhattacharya, K.; Glendening, J. M.; Stopa, E.; Vlassara, H.; Bucala, R.; Manogue, K.; Cerami, A. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 4766−4770. (26) Bush, A. I.; Pettingell, W. H.; Multhaup, G.; Paradis, M. D.; Vonsattel, J. P.; Gusella, J. F.; Beyreuther, K.; Masters, C. L.; Tanzi, R. E. Science 1994, 265, 1464−1467. (27) Stine, W. B., Jr.; Dahlgren, K. N.; Krafft, G. A.; LaDu, M. J. J. Biol. Chem. 2003, 278, 11612−11622. (28) Yamaguchi, T.; Yagi, H.; Goto, Y.; Matsuzaki, K.; Hoshino, M. Biochemistry 2010, 49, 7100−7107. (29) Goldsbury, C.; Baxa, U.; Simon, M. N.; Steven, A. C.; Engel, A.; Wall, J. S.; Aebi, U.; Muller, S. A. J. Struct. Biol. 2011, 173, 1−13. (30) Sachse, C.; Fandrich, M.; Grigorieff, N. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 7462−7466. (31) Lindgren, M.; Hammarstrom, P. FEBS J. 2010, 277, 1380−1388. (32) Wolfe, L. S.; Calabrese, M. F.; Nath, A.; Blaho, D. V.; Miranker, A. D.; Xiong, Y. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 16863−16868. (33) Hossain, S.; Grande, M.; Ahmadkhanov, G.; Pramanik, A. Exp. Mol. Pathol. 2007, 82, 169−174. (34) Baumketner, A.; Bernstein, S. L.; Wyttenbach, T.; Bitan, G.; Teplow, D. B.; Bowers, M. T.; Shea, J. E. Protein Sci. 2006, 15, 420− 428. (35) Grasso, G. Mass Spectrom. Rev. 2011, 30, 347−365. (36) Rohner, T. C.; Staab, D.; Stoeckli, M. Mech. Ageing Dev. 2005, 126, 177−185. (37) Bartolini, M.; Naldi, M.; Fiori, J.; Valle, F.; Biscarini, F.; Nicolau, D.; Andrisano, V. Anal. Biochem. 2011, 414, 215−225. (38) Yan, P.; Hu, X.; Song, H.; Yin, K.; Bateman, R. J.; Cirrito, J. R.; Xiao, Q.; Hsu, F. F.; Turk, J. W.; Xu, J.; Hsu, C. Y.; Holtzman, D. M.; Lee, J. M. J. Biol. Chem. 2006, 281, 24566−24574. (39) Grasso, G.; Rizzarelli, E.; Spoto, G. Biochim. Biophys. Acta, Proteins Proteomics 2008, 1784, 1122−1126. (40) Wu, C.; Murray, M.; Bernstein, S.; Condron, M.; Bitan, G.; Shea, J.; Bowers, M. J. Mol. Biol. 2009, 387 (2), 492−501. (41) Morgner, N.; Kleinschroth, T.; Barth, H. D.; Ludwig, B.; Brutschy, B. J. Am. Soc. Mass Spectrom. 2007, 18, 1429−1438. (42) Morgner, N.; Barth, H. D.; Brutschy, B. Aust. J. Chem. 2006, 59, 109−114. (43) Peters, I.; Igbavboa, U.; Schutt, T.; Haidari, S.; Hartig, U.; Rosello, X.; Bottner, S.; Copanaki, E.; Deller, T.; Kogel, D.; Wood, W. G.; Muller, W. E.; Eckert, G. P. Biochim. Biophys. Acta 2009, 1788, 964−72.

(44) Barrera, N. P.; Robinson, C. V. Annu. Rev. Biochem. 2011, 80, 247−271. (45) Soto, C.; Sigurdsson, E. M.; Morelli, L.; Kumar, R. A.; Castano, E. M.; Frangione, B. Nat. Med. 1998, 4, 822−826. (46) Schmidt, T. L.; Nandi, C. K.; Rasched, G.; Parui, P. P.; Brutschy, B.; Famulok, M.; Heckel, A. Angew. Chem., Int. Ed. Engl. 2007, 46, 4382−4384. (47) Tjernberg, L. O.; Pramanik, A.; Björling, S.; Thyberg, P.; Thyberg, J.; Nordstedt, C.; Berndt, K. D.; Terenius, L.; Rigler, R. Chem. Biol. 1998, 6, 53−62. (48) Rzepecki, P.; Nagel-Steger, L.; Feuerstein, S.; Linne, U.; Molt, O.; Zadmard, R.; Aschermann, K.; Wehner, M.; Schrader, T. J. Biol. Chem. 2004, 279, 47497−47505. (49) Wu, J. W.; Breydo, L.; Isas, J. M.; Lee, J.; Kuznetsov, Y. G.; Langen, R.; Glabe, C. J. Biol. Chem. 2010, 285, 6071−6079. (50) Austen, B. M.; Paleologou, K. E.; Ali, S. A. E.; Qureshi, M. M.; Allsop, D.; El-Agnaf, O. M. A. Biochemistry 2008, 47, 1984−1992. (51) Soto, C.; Kindy, M. S.; Baumann, M.; Frangione, B. Biochem. Biophys. Res. Commun. 1996, 226, 672−680. (52) Frydman-Marom, A.; Rechter, M.; Shefler, I; Bram, Y.; Shalev, D. E.; Gazit, E. Angew. Chem., Int. Ed. 2009, 48, 1981−1986. (53) Bitan, G.; Kirkitadze, M. D.; Lomakin, A.; Vollers, S. S.; Benedek, G. B.; Teplow, D. B. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 330−335. (54) Hoffmann, J.; Aslimovska, L.; Bamann, C.; Glaubitz, C.; Bamberg, E; Brutschy, B. Phys. Chem. Chem. Phys. 2010, 12, 3480− 3485. (55) Cole, R. B., Ed. Electrospray Ionization Mass Spectrometry : Fundamentals, Instrumentation, and Applications; John Wiley & Sons Inc.: New York, 1997. (56) Bitan, G.; Tarus, B.; Vollers, S. S.; Lashuel, H. A.; Condron, M. M.; Straub, J. E.; Teplow, D. B. J. Am. Chem. Soc. 2003, 125, 15359− 15365.

5284


Laser-Induced Liquid Bead Ion Desorption Mass Spectrometry: An

Recommend Documents