by Spontaneous Nonenzymatic Processes - ACS Publications

Article pubs.acs.org/ac

Amyloid Plaque in the Human Brain Can Decompose from Aβ(1-40/142) by Spontaneous Nonenzymatic Processes Brian Lyons,§,† Michael Friedrich,§ Mark Raftery,‡ and Roger Truscott*,§ §

Illawarra Health and Medical Research Institute, University of Wollongong, Northfields Avenue, Wollongong, New South Wales 2522, Australia † Save Sight Institute, Sydney Eye Hospital, University of Sydney, 8 Macquarie Street, Sydney, New South Wales 2001, Australia ‡ Biological Mass Spectrometry Facility, The University of New South Wales, Sydney, New South Wales 2052, Australia S Supporting Information *

ABSTRACT: The degradation of long-lived proteins in the body is an important aspect of aging, and much of the breakdown is due to the intrinsic instability of particular amino acids. In this study, peptides were examined to discover if spontaneous nonenzymatic reactions could be responsible for the composition of Alzheimer’s (AD) plaque in the human brain. The great majority of AD plaque consists of N-terminally truncated versions of Aβ(1-40/1-42), with the most abundant peptide commencing with Glu (residue 3 in Aβ1-40/1-42) that is present as pyroGlu. Several Asp residues are racemized in Aβ plaque, with residue 1 being predominantly LisoAsp and peptide bond cleavage next to Ser 8 is also evident. In peptides, loss of the two N-terminal amino acids as a diketopiperazine was demonstrated at pH 7. For the Aβ Nterminal hexapeptide, AspAlaGluPheArgHis, this resulted in the removal of AspAla diketopiperazine and the generation of Glu as the new N-terminal residue. The Glu cyclized readily to pyroGlu. This pathway was altered significantly by zinc, which promoted pyroGlu formation but decreased AspAla diketopiperazine release. Zinc also facilitated cleavage on the N-terminal side of Ser 8. Racemization of the original N-terminal Asp to L-isoAsp was also detected and loss of one amino acid from the N-terminus. These data are therefore entirely consistent with plaque in the human brain forming from deposition of Aβ(1-40/1-42) and, over time, decomposing spontaneously. Since amyloid plaque is present in the human brain for years prior to the onset of AD, gradual spontaneous changes to the polypeptides within it will alter its properties and those of the oligomers that can diffuse from it. Such incremental changes in composition may therefore contribute to the origin of AD-associated cytotoxicity.

A

Asp15 residues. These are spontaneous (i.e nonenzymatic) processes.16 Long-lived proteins are now known to be present in many tissues of the body, including the brain,17−19 and their decomposition over time may play a pivotal role in age-related human diseases.20,21 The origin of other major PTMs documented in plaque is less clear and could involve enzymes. Most of these PTMs were localized at the N-terminus of Aβ (Figure 1). For example, the single major peptide present in plaque has lost the two Nterminal amino acids, and the newly uncovered Glu residue is cyclized to pyroglutamate.12 In addition, in another abundant Aβ component, the N-terminal Asp has been converted to LisoAsp.22 The mechanisms of some PTMs, such as Asp isomerization, were investigated by Zirah et al. using mostly Nacetyl Aβ(1-10).23

myloid plaque in the brain is thought to play a key role in Alzheimers Disease (AD).1 The polypeptides in plaque are derived from proteolytic cleavage of Amyloid Precursor Protein (APP) to produce fragments Aβ(1-40) and Aβ(142)2−4 that then polymerize.5 A great deal of research on AD has been undertaken with a view to inhibiting the γ and β secretase enzymes involved in the proteolysis of APP and therefore reducing the flux of plaque precursors.6−8 Another strategy to treat AD has employed antibodies to target the peptides in plaque with the aim of clearing it from the brain. Chelation therapy, largely targeted at zinc and copper, also has potential for AD treatment.9,10 Thus, far, these approaches have not yielded an efficacious drug for human use. Surprisingly when plaque from the human brain was analyzed11−13 very little of it was found to be Aβ(1-40) or Aβ(1-42). Rather the polypeptides that composed senile and vascular plaque were found to be overwhelmingly derivatives of Aβ(1-40) and Aβ(1-42) that had undergone a number of posttranslational modifications (PTMs). Most of these PTMs were typical of long-lived proteins e.g. racemization of Ser14 and © 2016 American Chemical Society

Received: October 22, 2015 Accepted: February 4, 2016 Published: February 4, 2016 2675

DOI: 10.1021/acs.analchem.5b03891 Anal. Chem. 2016, 88, 2675−2684

Article

Analytical Chemistry

Figure 1. Major N-terminal modifications of Aβ (1-42) that have been documented in plaque from the human brain (11-13). Aside from isomerization of Asp 1, Aβ(2-42) and Aβ(3-42) are dominant species, with some evidence of Aβ(4-42). While Aβ(8-42) and Aβ(9-42) are also present, there is no evidence of Aβ(5-42), Aβ(6-42), or Aβ(7-42). The reasons for this distinctive pattern can be traced to spontaneous processes and appear to be governed by the properties of the particular amino acid residues at the Aβ N-terminus.

spontaneous processes. Furthermore, most reactions were significantly modulated by zinc, which is abundant in the brain. Our findings are consistent with the polypeptides in plaque being long-lived and undergoing little turnover.

Analysis of plaque revealed that the pattern of N-terminal amino acids is also quite characteristic. Although, as indicated above, loss of one and two amino acids from Aβ are major processes, no evidence for loss of four (Phe), five (Arg), or six (Asp) amino acids was found.12,24,25 If spontaneous chemical reactions are responsible for amino acid loss via “laddering”, this finding seems paradoxical. The origin of these PTMs is important to understand since they may affect the ability of therapeutic antibodies to interact with plaque. In addition, some of the Aβ forms appear to behave differently.26−29 For example, the pyroglutamate 3 form of Aβ accumulates selectively in lysosomes.30 PTMs of Aβ may also compromise the ability of enzymes, such as neprilysin,31 to degrade Aβ.32 Thus, although the enzymatic hydrolysis of APP leads initially to either Aβ(1-40)or (1-42),33 much less than 5% of plaque in the human brain consists of intact Aβ(1-40/142).11−13 Rather, Aβ plaque is characterized by the following PTMs: a) loss of two amino acids, Asp and Ala, from the Nterminus; b) cyclization of the resulting Glu N-terminal residue to pyroglutamate; c) “laddering” from the N-terminus involving the sequential removal of one amino acid. This process is selective since it yields Aβ polypeptides ending in Ala 2 and Glu 3 but not Arg 5, His 6, or Asp 7; d) isomerization of Asp residues, in particular the N-terminal Asp residue primarily to L-isoAsp and e) cleavage next to Ser 8. In addition to these diagnostic N-terminal PTMs, a number of other modifications further along the Aβ oligopeptide have been documented.34−36 In this study we show that all of the PTMs known to characterize the Aβ polypeptides in plaque can be explained by

■

MATERIALS AND METHODS

Peptide Incubations. All peptides were synthesized by GLS Biochem (Shanghai, China). All peptides were checked by HPLC and mass spectrometry/mass spectrometry (MS/MS) prior to use. Representative ESI MS/MS traces are in Figures S1−S3. Peptides were dissolved in triplicate (1 mg/mL pH 7.4) in 100 mM phosphate buffer or 50 mM Tris at 37 or 60 °C. 60 °C was used to accelerate reaction rates. For the 37 °C incubations, a drop of chloroform was added to each tube to prevent microbial growth. ZnCl2 was added to some incubations. Aliquots (20 μL) were analyzed by HPLC. A table describing all reaction conditions used is shown in Table S1. HPLC Analysis and Quantification. An Agilent 1100 HPLC system (Agilent Technologies, Santa Clara, CA, USA) controlled using Chemstation software and equipped with a PDA detector was used. Incubations were monitored at 280 and 216 nm. Separation of the peptides was achieved using a Jupiter Proteo 4 μm 90 Å column (150 mm × 4.6 mm ID) at 40 °C. The gradient was 0% B (0.1% TFA in acetonitrile) to 60% B (0.1% TFA in acetonitrile) over 15 min with a flow rate of 2 mL/min. A standard curve was generated for each peptide. The degree of modification was calculated based on the moles of each peptide formed as a percentage of moles of peptide present at the start of the incubation. The error bars refer to the standard deviation of three replicates. Analysis by MALDI MS/MS Mass Spectrometry. MALDI-MS analysis was performed using a Shimadzu (Kyoto, Japan) Axima TOF2 mass spectrometer used in 2676


Article


Figure 2. N-terminal modifications of DAEFRH. (a) HPLC trace following incubation of DAEFRH for 14 days at 60 °C. DA dkp absorbs at 216 nm to a smaller degree than the peptides and is shown expanded in the inset. Peak-X contains a mixture of (L-isoAsp)AEFRH, (D-Asp)AEFRH, and (DisoAsp)AEFRH, and these were characterized by capillary LC-MS in Figure 3. Time course for the formation of (b) DA dkp, (c) EFRH, (d) (pyroGlu)FRH, and (e) AFERH from DAEFRH. Peptides were incubated in 50 mM Tris, pH 7.4 ± 2 mM ZnCl2 as indicated.

■

reflectron positive ion mode. Peptides were prepared in αcyano-4-hydroxycinnamic acid (8 mg/mL) in 80% (v/v) acetonitrile, 0.1% (v/v) TFA. LC-MS/MS Analysis of Aβ(1-6). Isomeric peptides were separated by nano-LC using an Ultimate RSLC and autosampler system (Thermo Fisher Scientific, Amsterdam, Netherlands). Samples (0.5 μL, ∼10 fmol) were separated using a fritless nano column (75μ × ∼40 cm) containing C18 media (1.9 μ, 120 Å Dr Maisch, Ammerbuch-Entringen Germany) at 45 °C. Peptides were eluted using a linear gradient of buffer A to 15% buffer B (H2O:CH3CN 20:80, 0.1% formic acid) at 200 nL/min over 39 min. LC-MS/MS Analysis of Aβ(1-42). Peptides were resuspended in 50 μL of 3% (v/v) acetonitrile/0.1% (v/v) formic acid, briefly sonicated, and centrifuged at 16,000g for 5 min. Samples were separated by nano-LC using an Ultimate 3000 HPLC and autosampler system (Dionex, Amsterdam, Netherlands) coupled to an in-house built fritless nano 75 μm × 30 cm column packed with ReproSil Pur 120 C18 stationary phase (1.9 μm, Dr, Maisch GmbH, Germany). LC mobile phase buffers were comprised of A: 0.1% (v/v) formic acid and B: 80% (v/v) acetonitrile/0.1% (v/v) formic acid. Peptides were eluted using a linear gradient of 5% B to 40% B over 60 min and then 95% B wash over 1 min at a flow rate of 250 nL/min. The LC was coupled to a QExactive Plus Orbitrap mass spectrometer (Thermo Fisher Scientific, Scoresby). Column voltage was 2300 V, and the heated capillary was set to 275 °C. Positive ions were generated by electrospray, and the Orbitrap operated in data-dependent acquisition mode.

RESULTS The overall aim of this study was to determine if the above Nterminal alterations that take place to Aβ in the human brain could be reproduced in vitro. In order to do this, the Nterminus of Aβ was employed, since, as illustrated above, this is where the majority of peptide modifications have been characterized. To reduce complexity it was decided to primarily focus on the hexapeptide DAEFRH, corresponding to the first 6 amino acids of Aβ. This strategy avoided using peptides that contain other Asp sites e.g. Asp 7. Asp residues are well-known to racemize to four Asp isomers,37,38 and this would lead to the generation of very complex HPLC patterns and therefore make quantitative analysis difficult. Since the objective was to mimic physiological conditions, experiments were performed at pH 7, unless otherwise indicated. Use of Model Peptides. Prior to commencing these experiments and to demonstrate that the degradation of longer peptides, such as Aβ could validly be studied using the Nterminal hexapeptide of Aβ, it was important to show that amino acids further along the sequence did not exert a significant effect on reactions at the N-terminus. To do this, SP diketopiperazine (dkp) formation was examined from a series of peptides where residues 7 and 8 were altered (SPAVQSFTTIVE, SPAVQSFRRIVE, SPAVQSFKKIVE, and SPAVQSFEEIVE) (Figure S4a,b). These preliminary experiments demonstrated that dkp formation is a facile reaction at neutral pH that can contribute to peptide degradation. In addition, this process appeared to be independent of the peptide length and was unaffected by the nature of residues >7 amino acids removed from the Nterminus (Figure S4c). To confirm that dkp formation can occur at 37 °C, three peptides (SPSY, YPAATIPY, and SASY) 2677


Article


Figure 3. (a) Representative HPLC trace following incubation of AEFRH for 14 days. Time course for the formation of (b) EFRH and (c) (pyroGlu)FRH. (d) HPLC trace following incubation of EFRH for 7 days. (*) A small amount of FRH was also detected. Time course for the formation of (pyroGlu)FRH at (e) pH 7.4 and (f) pH 5. (g) Percentage of each Aβ peptide which remained following separate incubation for 7 days. (h) Ratio of (L-isoAsp)AEFRH to (L-Asp)AEFRH and (D-Asp)AEFRH + (D-isoAsp)AEFRH) to (L-Asp)AEFRH following incubation of DAEFRH. Peptides were incubated 60 °C ± 2 mM ZnCl2 in 50 mM Tris pH 7.4 or 50 mM MOPS pH 5 as indicated.

chain that can potentially interact with the α-amino group and alter its nucleophilicity. These initial experiments indicated that modifications of the N-terminus of a peptide, such as Aβ, could be studied using a truncated version and that the amino acid sequence at the Nterminus can greatly affect the outcome of spontaneous chemical reactions involving Aβ. Incubation of DAEFRH (Aβ1-6) - Diketopiperazine Formation. Aβ(1-6) was incubated and samples removed for analysis by HPLC. A representative HPLC trace is shown in Figure 2a. Characterization of the newly formed HPLC peaks by ESI and MALDI mass spectrometry and NMR spectroscopy together with comparisons with synthetic standards revealed that DA dkp formed (Figure 2b). If DA dkp generation was occurring, the other part of the peptide (i.e., EFRH) should be detected in approximately equal amounts, and this was indeed observed (Figure 2c). The incorporation of zinc into the incubation reduced the formation of DA dkp as well as EFRH. Incubation of DAEFRH (Aβ1-6) - Pyroglutamic Acid Formation. Since DA dkp was formed from (Aβ1-6), the stability of the remaining peptide portion (EFRH) was examined. It is known that peptides with N-terminal Gln or Glu can spontaneously cyclize to the pyroglutamyl form

were incubated at pH7. dkp formation occurred readily in SPSY and YPAATIPY (which contain penultimate Pro residues) with trace amounts observed for SASY (Figure S4d). This is consistent with Pro as the penultimate residue acting to promote dkp formation.39 Additional experiments suggested that the N-terminal amino acid sequence significantly affected dkp generation, and therefore this aspect was investigated further. As expected an N-terminal acetyl group prevented the loss of dkp (Figure S5a,b). Dkp formation was observed for all Pro 2 peptides (SPSY, SPAY, FPHSPSY) and was dependent on the amino acids present in positions 1, 2, and 3. Interestingly peptides with an N-terminal Pro underwent negligible dkp formation (PFHSPSY, PAHSPSY, PEHSPSY, and PDHSPSY), and of particular relevance to Aβ, two peptides with an N-terminal Ala (AAPSY and AAAPSY) failed to form a dkp. On this basis we would predict that if the Aβ peptide loses one amino acid (Asp) the resultant AlaGlu N-terminal peptide may not decompose via dkp formation. To add further weight to this hypothesis and its relevance to Aβ, two peptides with Glu as residue 2 (YEVRSDRDY and YEVR) were also intransigent (Figure S5cf). This may be due to the presence of a negatively charged side 2678


Article

Analytical Chemistry (pyroGlu).40 A time course showing the formation of (pyroGlu)FRH following incubation of Aβ(1-6) is shown in Figure 2d. Inclusion of zinc reduced the rate of (pyroGlu)FRH formation, possibly via zinc chelation of the alpha amino group as documented for dipeptides.41 Incubation of DAEFRH (Aβ1-6) - Loss of the N-Terminal Amino Acid. The loss of two amino acids at a time via dkp formation was anticipated based on the peptide incubations and from literature data.39 An unexpected cleavage that occurred in the Aβ(1-6) incubations was the loss of the Asp residue from the N-terminus of DAEFRH (Figure 2e). As the spontaneous loss of a single N-terminal residue does not appear to have been reported previously, this reaction was investigated in greater detail. Single amino acid loss was influenced by a number of factors including the nature and stereochemistry of the N-terminal residue (Figure S6a,b). Buffer concentration and trace metal ions also affected the rate (Figure S6c,d). In addition buffer influenced the reaction, with truncation occurring ∼5 times faster in phosphate (pH 7.4) compared to TES, HEPES, and Tris. The 2′ residue did not appear to have a major effect on the rate (Figure S6e), whereas an N-acetyl group prevented this cleavage suggesting a role for the α amino group. The most likely mechanism is intramolecular base-catalyzed hydrolysis involving the α-amino group. A characteristic of this reaction is that the rate occurs more slowly in D2O than in H2O.42 A peptide that readily loses the N-terminal Pro residue (PFHSPSY) was chosen to investigate this. PFHSPSY was incubated in phosphate buffer in D2O or H2O. The rate of Nterminal truncation was approximately half in D2O (4.6% truncated) compared to H2O (7.5% truncated) following 14 days incubation at 60 °C, supporting that single amino acid cleavage may be occurring via base catalyzed hydrolysis. Additional evidence for the involvement of the α-amino group came from the effect of pH on the rate of loss of Pro from PFHSPSY. When incubated at a variety of pHs, the initial rate correlated inversely with the degree of protonation of the α-amino group (pH5 0.44 pmol/h; pH5.4 0.66 pmol/h; pH6.4 1.16 pmol/h; pH7.4 01.94 pmol/h; pH8.4 2.91 pmol/h; pH9.4 5.29 pmol/h). Lastly, this reaction was found to occur at 37 °C (Figure S6f). Each of the peptide products (AEFRH, EFRH, (pyroGlu)FRH)) formed following the incubation of DAEFRH were then incubated separately in order to more fully establish the degradation pathway of Aβ. Incubation of AEFRH (Aβ2-6). When AEFRH (Aβ2-6) was incubated (Figure 3a), spontaneous loss of one amino acid was observed resulting in the formation of EFRH (Figure 3b). In addition (pyroGlu)FRH was also detected (Figure 3c). Zinc increased the rate of formation of EFRH and (pyroGlu)FRH. In line with our previous model peptide dkp data, where Ala (1) and Glu (2) peptides did not degrade by dkp formation, FRH was not observed, nor was AE dkp. A small degree of racemization of the N-terminal Ala residue was detected. A mechanism for this which involves nucleophilic attack of the alpha amino group has been described.43 Incubation of EFRH (Aβ3-6). To further examine the formation of pyroGlu, separate incubations of EFRH (Figure 2d) were performed, and the rate of conversion to (pyroGlu)FRH was measured at both pH 7.4 (Figure 3e) and pH 5 (Figure 3f). The data in Figure 3e were consistent with those seen in the incubation of Aβ(1-6) and Aβ(2-6) indicating that cyclization of EFRH occurs spontaneously at neutral pH and

does not depend on the presence of other components in the DAEFRH incubation mixture. When (pyroGlu)FRH was incubated separately, it showed no evidence for degradation irrespective of the addition of zinc (Figure S7a,b). A histogram summarizing the amount of each of the above Aβ peptides that remained following 7 days incubation is shown in Figure 3g. N-Terminal Isomerization of Asp. Isomerization of the N-terminal amino acid in peptides has been reported previously.43,44 Analysis of DAEFRH incubations by nanocapillary LC-MS/MS and comparison with synthetic standards enabled measurement of N-terminal isomerization. Under the chromatographic conditions used, the diastereomers DAEFRH and (L-isoAsp)AEFRH were completely resolved; however, (DAsp)AFERH and (D-isoAsp)AEFRH coeluted. For this reason, the ratio of (L-isoAsp)AEFRH to L-AspAEFRH and (DAsp)AFERH + (D-isoAsp)AEFRH to L-AspAEFRH were calculated at each time point (Figure 3h). This analysis revealed that after 28 days at 60 °C the L-isoAsp peptide was present in a ∼1:1 ratio with respect to starting peptide (LAspAEFRH). In addition, a smaller quantity of the (D-Asp)/(DisoAsp) peptide was formed. The fact that isoAsp versions were found shows that the N-terminal racemization of Asp in DAEFRH takes place via a succinimide intermediate.37 In a separate study (L-isoAsp)AEFRH was incubated under the same conditions as used for DAEFRH. No interconversion to the other Asp isomers was detected revealing that once the isoAsp version is formed, it is stable (Figure S7c,d). This stability of isoAsp forms of a peptide has been observed previously.45 Zinc did not have a measurable effect on the rate of N-terminal racemization. In order to demonstrate the physiological relevance of this modification, DAEFRH was also incubated at 37 °C. By 11 weeks 22% DAEFRH had converted to (L-isoAsp)AEFRH and ∼5% had converted to (D-Asp)AFERH/(D-isoAsp)AEFRH. Incubation of Ac-DAEFRHDSGY (N-Acetyl Aβ-10) -Truncation at Ser 8. Analysis of the N-terminal amino acids in Aβ plaque from the human brain showed clearly that while Asp 1, Ala 2, and Glu 3 peptides were abundant, peptides beginning with Arg 5, His 6, and Asp 7 were not.11−13 Mass spectral data suggest that Aβ(4-42) may also be present in the AD brain.46,47 On the other hand, significant quantities of Aβ peptides with N-terminal Ser 8 and Gly 9 were detected.48 These data suggested that a separate cleavage event at the Nterminus of Ser 8 could also be occurring in Aβ. If so, it would result in the generation of another set of N-terminal laddering reactions. To investigate truncation at the N-terminal side of Ser 8, Nacetyl Aβ(1-10) was synthesized and incubated ± ZnCl2 (Figure S8a,b). The addition of zinc resulted in ∼15% truncating after 2 weeks at 60 °C (Figure 4). Almost no truncation was observed in the zinc-free incubation. The involvement of the Ser hydroxyl group in peptide bond cleavage has been documented41,49,50 HPLC profiles of Nacetyl Aβ(1-10) following incubation also showed evidence for racemization of Asp7. This reaction was anticipated based on numerous other examples of Asp peptides13,22 including Aβ(116) 23 and was therefore not studied in more detail. Interestingly, when the same N-acetyl Aβ(1-10) peptide incorporating L-isoAsp at position 7 was incubated, it showed no evidence for formation of the other Asp isomers, indicating that like Asp 1, the isoAsp 7 is stable with little or no conversion to the other Asp isomers. 2679


Article


In the case of N-terminal isomerization in DAEFRH, conversion of L-Asp to L-isoAsp was the most prevalent reaction, and this finding is consistent with the analysis of authentic Aβ plaque.33 Lesser amounts of D-Asp/D-isoAsp were produced during the incubations. The presence of the various Asp isomeric forms is consistent with the involvement of a succinimide intermediate.37 The reason for the preferential formation of L-isoAsp presumably relates to the allowable conformations of the cyclic intermediate and/or the ability of this succinimide to racemize prior ring opening. Interestingly separate incubations of (L-isoAsp)AFERH revealed no evidence for conversion back to the other Asp isomeric forms. This is consistent with another study of isoAsp peptides45 and indicates that once the N-terminal Asp is racemized in Aβ, this version of the peptide may be stable. In agreement with other observations53 an isoAsp N-terminus may also be refractive to digestion by exopeptidases such as aminopeptidase. The major component of brain plaque is a peptide that has an N-terminal pyroglutamic acid corresponding to residue 3 of Aβ(1-40/1-42).12 It was demonstrated here that its formation can be explained by two spontaneous processes: first loss of two amino acids of Aβ as a DA diketopiperazine and second cyclization of the newly formed N-terminal Glu residue. Like Asp isomerization, modification of the N-terminus, such as pyroGlu formation, will hinder endopeptidases activity and should lead to stabilization of Aβ.54 Glutaminyl-peptide cyclotransferase (glutaminyl cyclase) catalyzes the conversion of N-terminal Gln residues of peptides to pyroGlu,55 and it can also catalyze the dehydration of N-terminal L-glutamyl residues. Involvement of an enzyme is not a requirement, however since spontaneous cyclization to pyroGlu occurred readily under our conditions. This finding will have a bearing on studies whose aim is inhibition of this enzyme as means to inhibit formation of the pyroGlu 3 form of Aβ. The pyroGlu form of Aβ has become of greater interest recently due to the discovery that it localizes in brain lysosomes30 where it contributes to dysfunction of these organelles and potentially to the function of neurons and glial cells. Cyclization of the N-terminal Glu residue also occurred at the pH of lysosomes (pH 5) (Figure 3e, 3f). Another process documented for the Aβ of brain plaque is “laddering”,3,56 which appears to involve the sequential loss of one amino acid at a time from the N-terminus. This too was observed in the incubation studies of DAEFRH (Figure 2e). Although it is another spontaneous reaction, the precise mechanism was unknown. A number of peptides, when incubated at pH 7, showed loss of the N-terminal amino acid (Figure S6), and a free α amino group was necessary. The rate was influenced by the type of amino acids present (Figure S6a), the stereochemistry of the N-terminal residue (Figure S6b), and the pH. The mechanism appears to entail attack of water on the terminal peptide bond with the hydrolysis being assisted by the amino group. EDTA (Figure S6d) reduced the rate suggesting that metal ions could also play a role. Our investigations suggest intramolecular base-catalyzed hydrolysis as the likely mechanism behind this process. Zinc markedly influenced the reaction pathways. Zinc is an abundant trace metal in the brain and has been implicated in AD.57−60 Others have shown that zinc at similar levels can affect peptide post-translational modification.23 With Aβ(1-6), zinc decreased dkp formation and the loss of one amino acid but did not significantly affect N-terminal Asp isomerization. In addition, zinc promoted cleavage on the N-terminal side of Ser

Figure 4. Zinc catalyzes cleavage of the peptide bond on the Nterminal side of Ser. Time course for truncation at Ser 8. AcDAEFRHDSGY was incubated in 50 mM Tris, pH 7.4 ± 2 mM ZnCl2.

Incubation of Aβ(1-42). To demonstrate that the modifications outlined above do indeed apply to full length Aβ, the synthetic peptide (Aβ1-42) was incubated and analyzed by ESI mass spectrometry. Unsurprisingly this revealed a mixture of modified peptides, but the loss of one amino acid from the N-terminus, loss of two amino acids, pyroGlu formation, and truncation at Ser 8 were readily observed (Table S2). Interestingly significant C-terminal laddering of peptides was observed, but this was found to be due mostly to the process of peptide synthesis since these were also present in the control Aβ(1-42) samples. The fact that the N-terminal modifications noted above were generally found in a number of the C-terminal laddered peptides adds weight to the proposal outlined earlier (Figure S4c) that these N-terminal processes occur independently of the amino acid sequence further along the polypeptide chain. Interestingly pyroGlu formation was observed predominantly in the later time point (Week-8) indicating that it may take longer to accumulate this modification, probably reflecting the fact that it is a multistep process.

■

DISCUSSION The results of this study show that the major processes documented for the peptides of Aβ plaque in the human brain are consistent with spontaneous reactions and can be replicated using model peptides. This extends a study23 by Zirah et al. where the majority of experiments were performed using an Nacetylated version of Aβ(1-10) and where it was shown that internal Asp residues, such as Asp 7, were racemized. Racemization of Asp residues in peptides has been observed previously, e.g. refs 15, 51, and 52, and was therefore not the focus of this investigation. The current study focused on the N-terminal region of Aβ where most characteristic PTM signatures of brain Aβ have been reported.11,12,33 Each of these processes i.e. racemization of the N-terminal Asp residue, loss of two amino acids, and cyclization of the resulting new N-terminal Glu residue to pyroglutamate were observed. In addition the loss of a single residue from the N-terminus was detected. This sequential loss of one amino acid at a time from plaque has been observed in vivo.48 Each of these reactions is discussed in greater detail below. 2680


Article


Figure 5. Intrinsic properties of the various intermediates involved in the spontaneous breakdown of Aβ determine the main N-termini of the ultimate products isolated from amyloid in the brain. Based on peptide studies, three main pathways are available to the N-terminal sequence (DAEFRH) of Aβ once it is cleaved from APP. i. Isomerization of the N-terminal Asp. If L-isoAsp forms, this version of Aβ is stable, with no reversion to the L-Asp or the other Asp isomers. ii. Loss of two amino acids as DA diketopiperazine. The newly produced Glu N-terminus then cyclizes to pyroGlu which is stable. iii. Loss of one amino acid to form AEFRH. This peptide does not readily lose AE diketopiperazine, and therefore the main pathway for further AEFRH degradation is loss of another single amino acid residue to form EFRH. The Glu N-terminus then cyclizes to pyroGlu with a small degree of conversion to FRH. The pyroGlu and L-isoAsp versions of Aβ are likely to be intransigent to exopeptidases. Red boxes indicate major N-terminal peptides present in plaque.

8 in Aβ(1-10). This finding may well account for the fact that significant amounts of Aβ in the brain has an N-terminal Ser.48 In other human proteins, age-related PTMs such as Asp and Ser racemization are localized to unstructured regions.14,61 This presumably reflects the need for conformational flexibility in order for spontaneous processes to occur. In accord with this prerequisite, several studies suggest that the N-terminus of Aβ is flexible.62−64 A similar study involving incubation of Aβ(1-16) was published recently by Zirah et al.23 The majority of experiments were performed using an N-acetylated version of this peptide. As shown in the current study, acetylating the α amino group of the peptide abolishes the characteristic reactions. This factor may have led to their conclusion that “the truncated molecular forms detected within fibrils presumably represent by-products of metabolic intermediates toward degradation and are not produced spontaneously upon protein aging” - page 2159. By contrast, our data demonstrate that most, if not all, reactions that characterize the Aβ peptides present in brain plaque can indeed be explained by spontaneous, nonenzymatic modifications. A key finding of the current investigations was that the amino acid sequence at the N-terminus of a peptide can markedly affect the products. The pathway of N-terminal degradation of Aβ is summarized schematically in Figure 5. It can be encapsulated at follows: Aβ(1-40/1-42) with the original Nterminal sequence DAEFRH is deposited following cleavage of APP. Three possible reactions can then take place: a. Isomerization of the N-terminal Asp. If L-isoAsp forms, this version of Aβ appears to be stable, with no interconversion to the other Asp isomers. The L-isoAsp version of Aβ is likely to be intransigent to exopeptidases. b. Loss of two amino acids as DA diketopiperazine. The newly produced Glu N-terminus then cyclizes to pyroGlu. c. Loss of one amino acid to form AEFRH. Due to the nature of the N-terminal and penultimate amino acids, this peptide does not readily lose AE diketopiperazine. Therefore, the main

pathway for further AEFRH degradation involves another loss of a single amino acid residue to form EFRH. The newly produced Glu N-terminus then cyclizes to pyroGlu as outlined in b. This pyroGlu is stable and is also likely to be intransigent to exopeptidases. These three features alone are sufficient to explain the analytical data on Aβ peptides isolated from brain plaque. In particular these data provide a rationale as to why little or no plaque commencing with Arg 6, His 6, and Asp 7 is detected. It is simply that the properties of the intermediates dictate the course of subsequent pathways. The finding that a significant amount of Aβ in plaque with N-terminal Ser 8 and Gly 948 may appear to be inconsistent with the hypothesis, however we demonstrated that zinc significantly affects Aβ breakdown. Specifically zinc catalyzes cleavage of the peptide bond on the N-terminal side of Ser. In doing so it uncovers a new Nterminal amino acid which can undergo another series of spontaneous reactions. What are some implications of these findings? First, it would appear that the major part of Aβ plaque in the brain is longlived. If there were rapid turnover of the peptides within plaque, coupled with excretion or digestion, it is not likely that the peptide profiles detected for bulk plaque56 would be as highly modified as they are by PTMs: all of which are consistent with slow spontaneous reactions. Second, that the composition of Aβ in brain plaque may change gradually over time. If the Aβ peptides within plaque are indeed long-lived, then the pattern of PTMs presumably increases over time. There is already some limited data to support this.29 It should be noted that a number of other modifications of Aβ such as isomerization of Asp 1,7 and Ser 8, 26 have been characterized.34 All of these also appear to be due to time-dependent spontaneous PTMs. Biochemical data indicate that oligomers are able to diffuse from amyloid in a process akin to equilibration,65 and there is a large amount of experimental information suggesting that these Aβ oligomers may be the main neurotoxic agents.66 Perhaps the strongest evidence that toxic oligomers do diffuse from amyloid 2681


Article


Figure 6. Proposed mechanisms that account for the range of N-terminal modifications present in Aβ from AD plaque (see Figure 1). These nonenzymatic N-terminal reactions can presumably take place in both Aβ(1-40) and Aβ(1-42). i. N-terminal Asp isomerizes predominantly to LisoAsp via a succinimide; ii, loss of one amino acid occurs via base catalyzed hydrolysis; iv, loss of two amino acids by dkp formation; v, Glu cyclizes spontaneously to pyroGlu; iii, the peptide bond adjacent to Ser 8 is hydrolyzed in a reaction catalyzed by zinc. Red text indicates the N-terminal residue.

■

and are important for AD comes from histological studies. It has been observed that a ring of dying neurons surrounds amyloid plaques in brain slices and that this region also is immunoreactive with antibodies to Aβ oligomers.67 One possible consequence arising from our data is that the oligomers that diffuse from brain plaque change with age. For example, it is conceivable that more highly modified Aβ oligomers present in older plaque are more toxic to neurons. In this sense it could be hypothesized that it may take time for amyloid plaque to become neurotoxic. Plaque in a normal (young) brain could be less of a problem for nerve cells than the “same” deposit after many years. This hypothesis provides a novel avenue for understanding the etiology of AD and would need to be tested experimentally. In summary, the data from peptide incubation experiments are entirely consistent with Aβ plaque in the human brain forming from the initial deposition of Aβ(1-40/1-42) and then decomposing via spontaneous processes. Mechanisms that could account for the range of nonenzymatic modifications of Aβ are summarized in Figure 6.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.5b03891. Figures S1−S3, representative ESI MS/MS traces; Figure S4, SP dkp formation examined from a series of peptides; Figure S5, N-terminal acetyl group prevented the loss of dkp; Figure S6, single amino acid loss influenced by a number of factors; Figure S7, (pyroGlu)FRH incubated separately, no interconversion to the other Asp isomers detected; Figure S8, zinc catalyzes cleavage of the peptide bond on the N-terminal side of Ser; Table S1, summary of the experimental conditions used for the various peptide incubations; Table S2, truncated peptides detected following incubation of Aβ(1-42) (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest. 2682


Article


■

(26) Russo, C.; Violani, E.; Salis, S.; Venezia, V.; Dolcini, V.; Damonte, G.; Benatti, U.; D’Arrigo, C.; Patrone, E.; Carlo, P.; Schettini, G. J. Neurochem. 2002, 82, 1480−1489. (27) Pike, C. J.; Overman, M. J.; Cotman, C. W. J. Biol. Chem. 1995, 270, 23895−23898. (28) Wirths, O.; Breyhan, H.; Cynis, H.; Schilling, S.; Demuth, H.-U.; Bayer, T. A. Acta Neuropathol. 2009, 118, 487−496. (29) Shin, Y.; Cho, H.; Fukumoto, H.; Shimizu, T.; Shirasawa, T.; Greenberg, S. M.; Rebeck, W. G. Acta Neuropathol 2003, 105, 252− 258. (30) De Kimpe, L.; van Haastert, E. S.; Kaminari, A.; Zwart, R.; Rutjes, H.; Hoozemans, J. J.; Scheper, W. Age 2013, 35, 673−687. (31) Wang, D.-S.; Iwata, N.; Hama, E.; Saido, T. C.; Dickson, D. W. Biochem. Biophys. Res. Commun. 2003, 310, 236−241. (32) Iwata, N.; Tsubuki, S.; Takaki, Y.; Shirotani, K.; Lu, B.; Gerard, N. P.; Gerard, C.; Hama, E.; Lee, H.-J.; Saido, T. C. Science 2001, 292, 1550−1552. (33) Iwatsubo, T.; Saido, T. C.; Mann, D. M.; Lee, V. M.; Trojanowski, J. Q. Am. J. Pathol. 1996, 149, 1823−1830. (34) Kummer, M. P.; Heneka, M. T. Alzheimer's Res. Ther. 2014, 6, 28. (35) Kubo, T.; Kumagae, Y.; Miller, C. A.; Kaneko, I. J. Neuropathol. Exp. Neurol. 2003, 62, 248−259. (36) Näslund, J.; Schierhorn, A.; Hellman, U.; Lannfelt, L.; Roses, A. D.; Tjernberg, L. O.; Silberring, J.; Gandy, S. E.; Winblad, B.; Greengard, P. Proc. Natl. Acad. Sci. U. S. A. 1994, 91, 8378−8382. (37) Stephenson, R. C.; Clarke, S. J. Biol. Chem. 1989, 264, 6164− 6170. (38) Yang, H.; Zubarev, R. A. Electrophoresis 2010, 31, 1764−1772. (39) Goolcharran, C.; Borchardt, R. T. J. Pharm. Sci. 1998, 87, 283− 288. (40) Chelius, D.; Jing, K.; Lueras, A.; Rehder, D. S.; Dillon, T. M.; Vizel, A.; Rajan, R. S.; Li, T.; Treuheit, M. J.; Bondarenko, P. V. Anal. Chem. 2006, 78, 2370−2376. (41) Yashiro, M.; Sonobe, Y.; Yamamura, A.; Takarada, T.; Komiyama, M.; Fujii, Y. Org. Biomol. Chem. 2003, 1, 629−632. (42) Fife, T. H.; Singh, R.; Bembi, R. J. Org. Chem. 2002, 67, 3179− 3183. (43) Lyons, B.; Kwan, A. H.; Jamie, J.; Truscott, R. J. FEBS J. 2013, 280, 1980−1990. (44) Smith, G. G.; Evans, R. C.; Biogeochemistry of Amino Acids; Wiley: 1980; pp 257−282. (45) Hooi, M. Y.; Raftery, M. J.; Truscott, R. J. Mech. Ageing Dev. 2013, 134, 103−109. (46) Antonios, G.; Saiepour, N.; Bouter, Y.; Richard, B. C.; Paetau, A.; Verkkoniemi-Ahola, A.; Lannfelt, L.; Ingelsson, M.; Kovacs, G. G.; Pillot, T. Acta Neuropathol. Commun. 2013, 1, 56. (47) Lewis, H.; Beher, D.; Cookson, N.; Oakley, A.; Piggott, M.; Morris, C.; Jaros, E.; Perry, R.; Ince, P.; Kenny, R. Neuropathol. Appl. Neurobiol. 2006, 32, 103−118. (48) Sergeant, N.; Bombois, S.; Ghestem, A.; Drobecq, H.; Kostanjevecki, V.; Missiaen, C.; Wattez, A.; David, J. P.; Vanmechelen, E.; Sergheraert, C.; Delacourte, A. J. Neurochem. 2003, 85, 1581−1591. (49) Lyons, B.; Jamie, J.; Truscott, R. W. Int. J. Pept. Res. Ther. 2011, 17, 131−135. (50) Lyons, B.; Jamie, J. F.; Truscott, R. J. Amino Acids 2014, 46, 199−207. (51) Fujii, N.; Momose, Y.; Harada, K. Int. J. Pept. Protein Res. 1996, 48, 118−122. (52) Kuge, K.; Brack, A.; Fujii, N. Chem. - Eur. J. 2007, 13, 5617− 5621. (53) Lyons, B.; Kwan, A. H.; Truscott, R. FEBS J. 2014, 281, 2945− 2955. (54) Liu, Y. D.; Goetze, A. M.; Bass, R. B.; Flynn, G. C. J. Biol. Chem. 2011, 286, 11211−11217. (55) Schilling, S.; Zeitschel, U.; Hoffmann, T.; Heiser, U.; Francke, M.; Kehlen, A.; Holzer, M.; Hutter-Paier, B.; Prokesch, M.; Windisch, M.; Jagla, W.; Schlenzig, D.; Lindner, C.; Rudolph, T.; Reuter, G.;

ACKNOWLEDGMENTS Funding for this work was provided by NHMRC # 1008667. The authors do not have any conflicts of interest to declare. LC-MS/MS results were obtained at the Bioanalytical Mass Spectrometry Facilities, University of New South Wales. Subsidized access to this facility is gratefully acknowledged.

■

REFERENCES

(1) Chiti, F.; Dobson, C. M. Annu. Rev. Biochem. 2006, 75, 333−366. (2) Kang, J.; Lemaire, H.-G.; Unterbeck, A.; Salbaum, J. M.; Masters, C. L.; Grzeschik, K.-H.; Multhaup, G.; Beyreuther, K.; Müller-Hill, B. Nature 1987, 325, 733−736. (3) Glenner, G.; Wong, C. Alzheimer Dis. Assoc. Disord. 1988, 2, 134. (4) Masters, C. L.; Selkoe, D. J. Cold Spring Harbor Perspect. Med. 2012, 2, a006262. (5) Ross, C. A.; Poirier, M. A. Nat. Med. 2004, 10 (Suppl), S10−17. (6) Walter, J.; Kaether, C.; Steiner, H.; Haass, C. Curr. Opin. Neurobiol. 2001, 11, 585−590. (7) Vassar, R.; Bennett, B. D.; Babu-Khan, S.; Kahn, S.; Mendiaz, E. A.; Denis, P.; Teplow, D. B.; Ross, S.; Amarante, P.; Loeloff, R.; Luo, Y.; Fisher, S.; Fuller, J.; Edenson, S.; Lile, J.; Jarosinski, M. A.; Biere, A. L.; Curran, E.; Burgess, T.; Louis, J.-C.; Collins, F.; Treanor, J.; Rogers, G.; Citron, M. Science 1999, 286, 735−741. (8) Yang, L. B.; Lindholm, K.; Yan, R.; Citron, M.; Xia, W.; Yang, X. L.; Beach, T.; Sue, L.; Wong, P.; Price, D.; Li, R.; Shen, Y. Nat. Med. 2003, 9, 3−4. (9) Hegde, M. L.; Bharathi, P.; Suram, A.; Venugopal, C.; Jagannathan, R.; Poddar, P.; Srinivas, P.; Sambamurti, K.; Rao, K. J.; Scancar, J.; Messori, L.; Zecca, L.; Zatta, P. J. Alzheimer’s Dis. 2009, 17, 457−468. (10) Cherny, R. A.; Atwood, C. S.; Xilinas, M. E.; Gray, D. N.; Jones, W. D.; McLean, C. A.; Barnham, K. J.; Volitakis, I.; Fraser, F. W.; Kim, Y.-S.; Huang, X.; Goldstein, L. E.; Moir, R. D.; Lim, J. T.; Beyreuther, K.; Zheng, H.; Tanzi, R. E.; Masters, C. L.; Bush, A. I. Neuron 2001, 30, 665−676. (11) Saido, T. C.; Iwatsubo, T.; Mann, D. M.; Shimada, H.; Ihara, Y.; Kawashima, S. Neuron 1995, 14, 457−466. (12) Saido, T. C.; Yamao-Harigaya, W.; Iwatsubo, T.; Kawashima, S. Neurosci. Lett. 1996, 215, 173−176. (13) Shimizu, T.; Watanabe, A.; Ogawara, M.; Mori, H.; Shirasawa, T. Arch. Biochem. Biophys. 2000, 381, 225−234. (14) Hooi, M. Y.; Raftery, M. J.; Truscott, R. J. Protein science 2013, 22, 93−100. (15) Hooi, M. Y. S.; Truscott, R. J. W. Age 2011, 33, 131−141. (16) Truscott, R. J. Rejuvenation Res. 2010, 13, 83−89. (17) Thayer, N. H.; Leverich, C. K.; Fitzgibbon, M. P.; Nelson, Z. W.; Henderson, K. A.; Gafken, P. R.; Hsu, J. J.; Gottschling, D. E. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 14019−14026. (18) Toyama, B. H.; Savas, J. N.; Park, S. K.; Harris, M. S.; Ingolia, N. T.; Yates, J. R.; Hetzer, M. W. Cell 2013, 154, 971−982. (19) D’Angelo, M. A.; Raices, M.; Panowski, S. H.; Hetzer, M. W. Cell 2009, 136, 284−295. (20) Truscott, R. J. W. Ageing Res. Rev. 2011, 10, 397−403. (21) Cloos, P.; Fledelius, C. Biochem. J. 2000, 345, 473−480. (22) Inoue, K.; Hosaka, D.; Mochizuki, N.; Akatsu, H.; Tsutsumiuchi, K.; Hashizume, Y.; Matsukawa, N.; Yamamoto, T.; Toyo’oka, T. Anal. Chem. 2014, 86, 797−804. (23) Zirah, S.; Kozin, S. A.; Mazur, A. K.; Blond, A.; Cheminant, M.; Segalas-Milazzo, I.; Debey, P.; Rebuffat, S. J. Biol. Chem. 2006, 281, 2151−2161. (24) Harigaya, Y.; Saido, T. C.; Eckman, C. B.; Prada, C.-M.; Shoji, M.; Younkin, S. G. Biochem. Biophys. Res. Commun. 2000, 276, 422− 427. (25) Wiltfang, J.; Esselmann, H.; Cupers, P.; Neumann, M.; Kretzschmar, H.; Beyermann, M.; Schleuder, D.; Jahn, H.; Rüther, E.; Kornhuber, J. J. Biol. Chem. 2001, 276, 42645−42657. 2683


Article

Analytical Chemistry Cynis, H.; Montag, D.; Demuth, H. U.; Rossner, S. Nat. Med. 2008, 14, 1106−1111. (56) Masters, C. L.; Simms, G.; Weinman, N. A.; Multhaup, G.; McDonald, B. L.; Beyreuther, K. Proc. Natl. Acad. Sci. U. S. A. 1985, 82, 4245−4249. (57) Deibel, M. A.; Ehmann, W. D.; Markesbery, W. R. J. Neurol. Sci. 1996, 143, 137−142. (58) Lovell, M. A.; Robertson, J. D.; Teesdale, W. J.; Campbell, J. L.; Markesbery, W. R. J. Neurol. Sci. 1998, 158, 47−52. (59) Markesbery, W. R. Free Radical Biol. Med. 1997, 23, 134−147. (60) Schrag, M.; Mueller, C.; Oyoyo, U.; Smith, M. A.; Kirsch, W. M. Prog. Neurobiol. 2011, 94, 296−306. (61) Hooi, M. Y.; Raftery, M. J.; Truscott, R. J. Protein science 2012, 21, 1074−1079. (62) Meral, D.; Urbanc, B. J. Mol. Biol. 2013, 425, 2260−2275. (63) Hubin, E.; van Nuland, N. A.; Broersen, K.; Pauwels, K. Cell. Mol. Life Sci. 2014, 71, 3507−3521. (64) Zhang, A.; Qi, W.; Good, T. A.; Fernandez, E. J. Biophys. J. 2009, 96, 1091−1104. (65) Maggio, J. E.; Stimson, E. R.; Ghilardi, J. R.; Allen, C. J.; Dahl, C. E.; Whitcomb, D. C.; Vigna, S. R.; Vinters, H. V.; Labenski, M. E.; Mantyh, P. W. Proc. Natl. Acad. Sci. U. S. A. 1992, 89, 5462−5466. (66) Benilova, I.; Karran, E.; De Strooper, B. Nat. Neurosci. 2012, 15, 349−357. (67) Meyer-Luehmann, M.; Spires-Jones, T. L.; Prada, C.; GarciaAlloza, M.; de Calignon, A.; Rozkalne, A.; Koenigsknecht-Talboo, J.; Holtzman, D. M.; Bacskai, B. J.; Hyman, B. T. Nature 2008, 451, 720− 724.

2684


by Spontaneous Nonenzymatic Processes - ACS Publications

Recommend Documents