Physiochemical characterization of the ... - Wiley Online Library

Physiochemical characterization of the Alzheimer’s disease-related peptides Ab1–42Arctic and Ab1–42wt Ann-Sofi Johansson1, Fredrik Berglind-Dehlin1, Go¨ran Karlsson2, Katarina Edwards2, Pa¨r Gellerfors1 and Lars Lannfelt1 1 Department of Public Health and Caring Sciences, Uppsala University, Rudbeck Laboratory, Sweden 2 Department of Physical Chemistry, Uppsala University, Sweden

Keywords amyloid-b; Arctic; fibrillization; oligomerization; protofibrils Correspondence A.-S. Johansson, Department of Public Health and Caring Sciences, Rudbeck Laboratory, Dag Hammarskjolds vag 20, SE-75185 Uppsala, Sweden Fax: +46 18 4714808 Tel: +46 18 4715030 E-mail: [email protected] (Received 19 January 2006, revised 5 April 2006, accepted 10 April 2006) doi:10.1111/j.1742-4658.2006.05263.x

The amyloid b peptide (Ab) is crucial for the pathogenesis of Alzheimer’s disease. Aggregation of monomeric Ab into insoluble amyloid fibrils proceeds through several soluble Ab intermediates, including protofibrils, which are believed to be central in the disease process. The main reason for this is their implication in familial Alzheimer’s disease with the Arctic amyloid precursor protein mutation (E693G). This mutation gives rise to early onset Alzheimer’s disease, and synthetic Ab1–40Arctic displays an enhanced rate of protofibril formation in vitro [Nilsberth C, Westlind-Danielsson A, Eckman CB, Condron MM, Axelman K, Forsell C, Stenh C, Luthman J, Teplow DB, Younkin SG, Naslund J & Lannfelt L. (2001) Nat Neurosci 4, 887–893]. To increase our understanding of the mechanisms involved in Ab aggregation, especially Ab monomer oligomerization into protofibrils and protofibril fibrillization into fibrils, the kinetics of Ab1– 42wt and Ab1–42Arctic aggregation were examined under different physiochemical conditions, such as concentration, temperature, ionic strength and pH. We used size exclusion chromatography for this purpose, where monomers are separated from protofibrils, and fibrils are separated from protofibrils in a centrifugation step. The Arctic mutation significantly accelerated both Ab1–42wt protofibril formation and protofibril fibrillization. In addition, we demonstrated that two distinct chemical processes – monomer oligomerization and protofibril fibrillization – were affected differently by changes in the micro-environment and that the Arctic mutation alters the peptide response to such changes.

Alzheimer’s disease (AD) is the main cause of dementia, affecting 60–70% of all dementia cases [1]. One of the pathological hallmarks of the disease is neuritic plaques located in the limbic and association cortex in the brain. The plaques are mainly composed of the 40–42 amino acid-long amyloid b peptide (Ab) in fibrillar form, produced by proteolytic processing of the amyloid precursor protein (APP). The aggregation of monomeric Ab into insoluble fibrils proceeds through several soluble Ab intermediates (dimer, trimer, oligomer), including protofibrils. The properties

of these intermediates, their neurotoxicity and their steady-state level in vivo in the brain of patients with AD are not known. However, in vitro data has shown that one of these intermediates, the protofibril, is formed in significant quantities during the aggregation of Ab [2,3], and the protofibril has been suggested to be a central pathogen in AD [4]. Protofibrils are soluble intermediates with a diameter of 6–8 nm and a beaded appearance in electron microscopy (EM) (for review, see ref [5]). They have been shown to be neurotoxic [6] and to alter neuronal electrophysiological

Abbreviations AD, Alzheimer’s disease; Ab, amyloid-b peptide; APP, amyloid precursor protein; Arc, Arctic; Cryo-TEM, Cryo transmission electron microscopy; EA, activation energy; EM, electron microscopy; SEC, size exclusion chromatography.

2618

FEBS Journal 273 (2006) 2618–2630 ª 2006 The Authors Journal compilation ª 2006 FEBS

A.-S. Johansson et al.

parameters [7]. Protofibrils have also been implicated in familial AD, as the Arctic (Arc) APP (E693G) mutation, AbE22G, gives rise to early onset AD. Ab1– 40Arc displayed an enhanced rate of protofibril formation in vitro [8] and completely blocked long-term potentiation in rats in vivo [9]. Furthermore, the Dutch APP mutation, AbE22Q, resulting in cerebral hemorrhage and AD-like symptoms, has been shown to affect the aggregation of the Ab1–40 peptide, with enhanced protofibril formation [2]. Ab1–42wt also form protofibrils to a higher extent than Ab1–40wt [2,3], which is interesting as most AD mutations increase Ab1–42wt levels [10]. Protofibrils are also formed from proteins implicated in other neurodegenerative diseases, such as familial amyloidotic polyneuropathy [11], Parkinson’s disease [12,13], familial British dementia [14] and Huntington’s disease [15], indicating a common mechanism in neurodegenerative diseases. To increase further the understanding of the mechanisms in Ab aggregation, especially Ab monomer oligomerization to protofibrils and protofibril fibrillization to insoluble fibrils, the kinetics of Ab1–42wt and Ab1– 42Arc fibrillization were examined under the following different physiochemical conditions: Ab concentration; temperature; ionic strength; and pH. Size exclusion chromatography (SEC) was used to monitor monomer oligomerization into protofibrils and protofibril fibrillization into fibrils over time. Monomer oligomerization was defined as the decay in monomer peak area. The monomer peak is also suggested to contain dimers [2], but will herein, for simplicity, be referred to as monomer. The Ab species eluting in the void volume of the column, but not pelleted in a centrifugation step, are defined as protofibrils. Pelleted species are defined as fibrils (i.e. loss of the protofibril area corresponds to the amount of fibrils formed). This is the first extensive study of how an Ab mutation influences the aggregation kinetics in response to physiochemical changes. The results demonstrate that the two distinct processes – monomer oligomerization and protofibril fibrillization – are affected differently by changes in the micro-environment and also that the Arctic mutation significantly accelerates the aggregation process.

Results Ab1–42Arc is a very hydrophobic peptide with a high tendency to aggregate, leading to adsorption to the column matrix and variability between experiments. To overcome this technical problem, the conditions for SEC had to be modified. Different detergents were tested for their capacity to reduce adsorption of Ab1–

Physiochemical characterization of Ab1–42Arctic

Fig. 1. Reduced protofibril area is accompanied by an increase in Thioflavin T (ThT) positive pelleted material. Ab1–42Arc (50 lM) was incubated at 37 C. Before analysis, Tween-20 was added and the sample was centrifuged at 17 900 g for 5 min. The amount of soluble amyloid-b peptide (Ab) in the supernatants was analyzed using size exclusion chromatography (SEC) and the amount of Ab in pellets was analyzed with amino acid analysis. Binding of the pelleted material to the fluorescent dye ThT was also evaluated. Supernatants eluted as one peak in the void volume of the SEC column. The amount soluble Ab left after incubation and centrifugation was calculated from the area of the void peak with the assumption that the area at time zero corresponds to the total amount of Ab in the sample. , nmol Ab in the supernatant determined by peak area from SEC; m, nmol Ab in pelleted material determined by amino acid analysis; h, ThT fluorescence of pelleted material.

42Arc to the column matrix. Tween-20 was found to minimize this interaction. Recovery of total Ab (the sum of protofibril and monomer areas) from the column was 70% under these conditions, as determined by injection of Ab with and without the column. Tween-20 does affect the aggregation process slightly by accelerating monomer oligomerization and stabilizing protofibrils (data not shown). However, Tween20 was present only during chromatography, not during the aggregation process, to minimize interference. The amount of Ab lost as a result of centrifugation was recovered in the pellet, as measured by amino acid analysis. Moreover, Thioflavin T (ThT) fluorescence of the pelleted material increased over time (Fig. 1). To ensure complete dissolution of the peptide and to avoid pre-existing aggregates, the peptides were dissolved in dilute sodium hydroxide since Ab dissolves very well at basic pH. As the pH is adjusted to physiological pH, the iso-electric point (pI ¼ 5.3 for Ab1– 42wt) where proteins are less soluble is avoided. Ab1–42Arc assembles into both protofibrils and fibrils faster than the wild-type monomer Figure 2B,C shows typical SEC chromatograms of Ab1–42wt and Ab1–42Arc, immediately after the


2619


Monomer oligomerization

A Monomer

Intensity (AU)

B


Protofibril fibrillization Protofibril

Fibril

0.20

0.15

0.10

0.05

0.00

Intensity (AU)

C

0

5

10

15

20

0

5

10

15

20

25

30

35

30

35

0.20

0.15

0.10

0.05

0.00 25

Retention time (min)

D

Fig. 2. The Arctic (Arc) mutation accelerates monomer oligomerization and protofibril fibrillization. (A) The amyloid-b peptide (Ab) aggregation process. The aggregation of monomers (i.e. decay in the monomer peak area) is denoted monomer oligomerization. Aggregation of soluble protofibrils into insoluble fibrils (i.e. decay in the protofibril peak area) is denoted protofibril fibrillization. (B,C) The Arc mutation accelerates Ab1–42 protofibril formation. Size exclusion chromatography (SEC) analysis of freshly dissolved Ab1– 42wt (B) and Ab1–42Arc (C), displaying the accelerated protofibril formation for Ab1–42Arc. Peptides were dissolved in NaOH ⁄ NaCl ⁄ Pi to a final concentration of 50 lM, and the monomer and protofibril content was measured using SEC. Protofibrils elute at 12 min and monomers at 20 min. (D) Kinetics of Ab1–42wt monomer oligomerization and Ab1–42wt and Ab1–42Arc protofibril fibrillization. Ab was incubated under standard conditions (37 C, pH 7.8, 0.1 M NaCl, 50 lM Ab) and assayed for Ab monomer and protofibril content as a function of time, using SEC. Each data value is the mean of 12 experiments. Error bars represent the standard deviation of the mean. Half lives under these experimental conditions were estimated by fitting the data to a one-phase exponential decay function. Based on this curve fit, the Ab1–42Arc protofibril fibrillization half life was estimated to be 36 min (SEM 6 min, n ¼ 12), the Ab1–42wt protofibril fibrillization half life to 7 h 32 min (SEM 58 min, n ¼ 12)], and the half life for Ab1–42wt monomer oligomerization to 2 h 18 min (SEM 18.8 min, n ¼ 12).

2620

peptides have been dissolved in NaOH and neutralized in NaCl ⁄ Pi. Approximately 50% of the Ab1–42wt monomer is converted into protofibrils eluting at 12 min, and the remaining Ab1–42wt remain as monomeric Ab1–42wt, eluting at 20 min. Interestingly, the Ab1–42Arc peptide was almost entirely (95%) converted to protofibrils under the same conditions, with only a minor (5%) monomeric peak remaining, demonstrating the strong protofibril formation properties conferred by the Arctic mutation. Very small amounts of Ab1–42wt and Ab1–42Arc intermediates, eluting between the protofibril peak and the monomeric peak were detected as a slight elevation of the baselines, without any distinct peaks. However, these Ab intermediates are not considered in the kinetic studies. The SEC method was used to determine the Ab monomer oligomerization and protofibril fibrillization kinetics (Fig. 2D). To study the kinetics under physiological conditions (pH 7.8 and 37 C), Ab1–42wt and Ab1–42Arc were incubated for different time periods and assayed for changes in Ab protofibril and monomer contents. The half life for Ab1–42Arc protofibril fibrillization under these conditions was estimated to be 30 min, more than 10-fold shorter than the corresponding half life for Ab1–42wt, showing the high protofibril fibrillization rate of Ab1–42Arc (Fig. 2D). The half life for Ab1–42wt monomer oligomerization was determined to be 2 h, significantly shorter than the Ab1–42wt protofibril fibrillization rate. A half life value for Ab1–42Arc monomer oligomerization could not be determined, as a result of the high oligomerization rate for Ab1–42Arc. Ab1–42wt monomer oligomerization is a more energy-dependent process than protofibril fibrillization The Ab aggregation process was found to be highly temperature dependent (Fig. 3). The half life for protofibril fibrillization was significantly longer at 5 C than at 25 C or 37 C, for both Ab1–42wt and Ab1– 42Arc. Also, the Ab1–42wt monomer oligomerization was very dependent on temperature, even though it did not reach significance as a result of the high variation in monomer oligomerization rate at 5 C. The data displayed a linear relationship in an Arrhenius plot where the natural logarithm of the estimated rate constants (k) were plotted against the inverse temperature (Fig. 4). From the slopes, activation energies were calculated. The Ab1–42wt monomer oligomerization activation energy (EA) was estimated to be approximately twofold higher than the EA of




A

B

C

D

Fig. 3. Temperature dependence of (A) Ab1–42wt protofibril fibrillization, (B) Ab1–42Arc protofibril fibrillization and (C) Ab1–42wt monomer oligomerization. Amyloid-b peptide (Ab) was incubated at 5 C, 25 C and 37 C, and assayed for Ab monomer and protofibril content as a function of time, using size exclusion chromatography (SEC). Other parameters were set to standard conditions (pH 7.8, 0.1 M NaCl, 50 lM Ab). Each data value is the mean of three experiments. Error bars represent the standard deviation of the mean. (D) Protofibril and monomer half lives were estimated from a one-phase exponential decay curve fit of data in A, B and C. Statistical significance was analyzed using analysis of variance (ANOVA). *P < 0.05.

Ab1–42wt protofibril fibrillization. The EA for Ab1– 42Arc protofibril fibrillization was slightly lower than the EA for wild-type protofibril fibrillization. The effect of Ab peptide concentration on the aggregation process Aggregation kinetics of Ab1–42wt and Ab1–42Arc were studied in relation to peptide concentration. Ab concentrations of 75, 50 and 25 lm were examined. No significant effect on either the rate of Ab1–42wt monomer oligomerization or the protofibril fibrillization rate was found (Fig. 5D). However, the initial and maximum amount of protofibrils formed for Ab1–42wt increased with increasing concentration (Fig. 5A). The Ab1–42Arc protofibril fibrillization rate was found to be decreased twofold at 75 lm compared with 50 and 25 lm (Fig. 5D), and the initial amount of Ab1–42Arc protofibrils was elevated at 75 lm (Fig. 5B).

Ab1–42Arc protofibril fibrillization is more dependent on ionic strength than Ab1–42wt The aggregation kinetics of Ab1–42wt and Ab1–42Arc were studied in relation to NaCl concentrations of 0.5, 0.1 and 0 m. The Ab1–42wt monomer oligomerization rate increased with increasing NaCl concentration (Fig. 6C,D) with a significantly shorter half life at 0.1 m NaCl compared with no NaCl. In contrast, Ab1– 42wt protofibril fibrillization did not display a clear dependence on ionic strength, even though the half life for the highest salt concentration was reduced. Interestingly, the effect on Ab1–42Arc protofibril fibrillization was more pronounced, with a significantly reduced half time with increasing salt concentration (Fig. 6D). Ab1–42Arc aggregates rapidly at high pH The aggregation kinetics of Ab1–42wt and Ab1–42Arc were studied at pH 5.2, 7.8 and 9.1. Ab1–42wt mono-


2621



wild-type protofibrils, but otherwise similar (Fig. 8B). However, Ab1–42Arc protofibrils commonly assembled in large, dense clusters (inset, Fig. 8B). These appeared to be different from the clusters observed in the Ab1– 42Arc fibril preparations. The large clusters in the Ab1–42Arc fibril preparations contained a number of long stretched structures, whereas the Ab1–42Arc protofibril clusters contained smaller structures adhering to each other.

Discussion

Fig. 4. Activation energies of Ab1–42wt monomer oligomerization and Ab1–42wt and Ab1–42Arc protofibril fibrillization. The natural logarithm of the rate constants for Ab1–42wt monomer oligomerization (.), Ab1–42wt protofibril fibrillization (n) and Ab1–42Arc protofibril fibrillization (d), were plotted as an inverse function of temperature, according to the Arrhenius law. Rate constants were derived from the half lives from the curves fitted to a one-phase exponential decay function (Fig. 3D). Activation energies (EA) can be derived from the slope of the linear curves. EA for Ab1–42wt monomer oligomerization was calculated to 115 ± 3 kJÆmole)1, for Ab1–42wt protofibril fibrillization to 50 ± 5 kJÆmole)1 and for Ab1– 42Arc protofibril fibrillization to 43 kJÆmole)1.

mer oligomerization kinetics were clearly dependent on pH, with an increased rate at decreasing pH (Fig. 7C). No half life value could be derived for pH 5.2, as hardly any monomers could be detected. The Ab1– 42Arc protofibril fibrillization behaved in a similar manner, with increasing fibrillization rate with decreasing pH, even though the difference between pH 7.8 and 9.1 was not significant. The most striking difference between pH 7.8 and 9.1 was the small amount of protofibrils formed for Ab1–42wt at pH 9.1, whereas Ab1–42Arc formed the same amount of protofibrils at pH 9.1 as at pH 7.8 (Fig. 7A,B). Cryo-transmission electron microscopy Fibrillar Ab1–42wt and Ab1–42Arc were analyzed with cryo-transmission electron microscopy (cryoTEM). Fibrillar Ab1–42wt preparations contained long stretched fibers, several hundreds of nanometers long (Fig. 8C). Fibrillar Ab1–42Arc was generally more compact and somewhat amorphous, with fibrils adhering to each other in large clusters (Fig. 8A). Ab1–42wt protofibril preparations were different from Ab1–42wt fibrils displaying typically small curly or globular structures (Fig. 8D), < 50 nm long. Ab1– 42Arc protofibrils seemed to be somewhat larger than 2622

These studies demonstrate that aggregation of Ab1– 42Arc and Ab1–42wt into fibrils proceeds via a soluble population of intermediates (i.e. protofibrils) eluting in the void volume of the SEC column. Other possible intermediates are likely to be unstable with low steadystate levels. We have studied how the Ab aggregation process is altered when the Arctic mutation is present in the amyloidogenic peptide, Ab1–42, as a way to receive further insight into the disease process. We demonstrate that protofibril formation and fibrillization is highly accelerated for the Arctic Ab1–42 peptide, as compared with Ab1–42wt. In addition, we illustrate that the early and late aggregation process for the two peptides are influenced to different degrees in response to physiochemical changes in the microenvironment. Morphology of protofibrils and fibrils For these studies, we used size exclusion chromatography to assay monomers and protofibrils as a function of Ab incubation time. Fibril formation is measured indirectly, by the decline in protofibril area. As the protofibrils reach a size where they fall out of solution, they are pelleted in a centrifugation step. These pellets are ThT positive and are therefore defined as fibrils. The fibrillar nature of pelleted Ab1–42wt species was confirmed by cryo-TEM, where long stretched fibrils were observed. The somewhat more amorphous appearance of Ab1–42Arc fibrils has previously been described for Ab1–40Arc fibrils [16]. The Ab1–42Arc fibrils seem to have a high propensity to adhere to each other, resulting in large fibrillar clusters. Ab1–42Arc protofibrils also adhered to each other, forming large clusters containing small structures, but separated protofibrils were not uncommon in the sample. Ab1–42wt and Ab1–42Arc protofibrils were much smaller than fibrils, and did not contrast as well as fibrils. These images were therefore slightly underfocused to increase the contrast. The Arctic protofibrils seemed to be larger than wild-type protofibrils, in




A

B

C

D

Fig. 5. Concentration dependence of (A) Ab1–42wt protofibril fibrillization (B) Ab1–42Arc protofibril fibrillization and (C) Ab1–42wt monomer oligomerization. A 100 lM concentration of amyloid-b peptide (Ab) was diluted to 75 lM, 50 lM and 25 lM in NaCl ⁄ Pi and assayed for Ab monomer and protofibril content as a function of time, using size exclusion chromatography (SEC). Other parameters were set to standard conditions (37 C, pH 7.8, 0.1 M NaCl). All peak areas were normalized to 50 lM. Each data value is the mean of three experiments. Error bars represent the standard deviation of the mean. (D) Protofibril and monomer half lives were estimated from a one-phase exponential decay curve fit of data in A, B and C. Statistical significance was analyzed with ANOVA. *P < 0.05.

agreement with the larger Ab1–40Arc fibrils described by Nilsberth et al. [8]. By using cryo-TEM instead of the more common negative staining technique, structures can be visualized without harsh staining techniques which may disrupt or induce oligomeric structures. Also, artifacts from the staining are completely avoided. The Arctic mutation The observed accelerated protofibril formation for Ab1–42Arc compared with wild type is in agreement with the results of Nilsberth et al. [8], where an increased protofibril formation was reported for Ab1–40Arc. Here, we also report an increased fibrillogenesis for Ab1–42Arc, in addition to the accelerated protofibril formation. This has not previously been observed for Ab1–40Arc [8], underlining the importance of the two hydrophobic C-terminal amino acids (Ile, Ala) in the fibrillization process. The fast aggregation of Ab1–42Arc is also consistent

with the data of Lashuel et al. [17], where preliminary studies show an increased formation of protofibrils for Ab1–42Arc compared with Ab1–40Arc, as measured by EM. The dramatic effect of the Arctic mutation on aggregation rate could be a result of the loss of charge, resulting from the substitution of the negatively charged glutamic acid for a glycine. Several mutations with increased aggregation rate have a loss of charge in this region. Besides the Arctic mutation, also the Dutch mutation (E22Q) [2], Iowa mutation (D23N) [18] and the nondisease-related mutations, E22A and E22V [19], accelerate Ab aggregation and have a loss of charge. Temperature and activation energies The fibrillization rate of Ab was found to be dependent on temperature, in agreement with previous studies [20–24]. An increase in temperature accelerated the aggregation rate, in agreement with CD studies, where higher temperature increased the b-structure content of


2623



A

B

C

D

Fig. 6. Ion strength dependence of (A) Ab1–42wt protofibril fibrillization, (B) Ab1–42Arc protofibril fibrillization and (C) Ab1–42wt monomer oligomerization. Amyloid-b peptide (Ab) was incubated in the presence of 0.5 M NaCl, 0.1 M NaCl or no NaCl, and assayed for Ab monomer and protofibril content as a function of time, using size exclusion chromatography (SEC). Other parameters were set to standard conditions (37 C, pH 7.8, 50 lM Ab). Each data value is the mean of three experiments. Error bars represent the standard deviation of the mean. (D) Protofibril and monomer half lives were estimated from a one-phase exponential decay curve fit of data in A, B and C. Statistical significance was analyzed with ANOVA. *P < 0.05. (1) The bar represents the half life derived from the curve fit from the three experiments together.

Ab1–40 in aqueous solution [25] and decreased temperatures shifted the structure of Ab12–28 towards a left-handed 31 helix structure [26]. According to the Arrhenius equation:

Ionic strength

ln k2 =k1 ¼ ðEA =RÞð1=T1 1=T2 Þ; the activation energy (EA) of a reaction is dependent on temperature (T) and the rate constant (k). We used this equation to estimate the activation energy for monomer oligomerization into protofibrils and protofibril fibrillization into insoluble fibrils. The highest activation energy was found for Ab1–42wt monomer oligomerization, with a estimated value of 115 kJÆ mol)1, reflecting the high temperature dependence for this process. This can be compared to the value of 96 kJÆmol)1 reported by Kusumoto et al. for Ab1–40 fibril elongation, measured by quasi-elastic light scattering [24]. The process measured was interpreted as association of monomers to fibril ends. The high activation energy for Ab1–42wt monomer oligomerization 2624

suggests that the early oligomerization process involves conformational changes to a higher degree than protofibril fibrillization.

Monomer oligomerization was also highly dependent on ionic strength, indicating hydrophobic interactions. The maximum amount of protofibrils formed was increased with increasing NaCl concentration, indicating that NaCl does not affect the protofibril fibrillization process to the same degree as monomer oligomerization, resulting in accumulation of protofibrils. On the contrary, Arctic protofibril fibrillization is clearly dependent on the ionic strength, suggesting a higher degree of hydrophobic interaction for Arctic protofibril fibrillization than wild-type protofibril fibrillization. The fibrillization rate of Ab1–40 protofibrils has previously been shown to be accelerated by an increased concentration of NaCl [20], demonstrating




A

B

C

D

Fig. 7. pH dependence of (A) Ab1–42wt protofibril fibrillization, (B) Ab1–42Arc protofibril fibrillization and (C) Ab1–42wt monomer oligomerization. Amyloid-b peptide (Ab) was incubated at pH 5.2, pH 7.8 or pH 9.1 and assayed for Ab monomer and protofibril content as a function of time, using size exclusion chromatography (SEC). Other parameters were set to standard conditions (37 C, 0.1 M NaCl, 50 lM Ab). Each data value is the mean of three experiments. Error bars represent the standard deviation of the mean. (D) Protofibril and monomer half lives were estimated from a one-phase exponential decay curve fit of data in A, B and C. Statistical significance was analyzed with ANOVA. *P < 0.05. (1) The bar represents the half life derived from the curve fit from the three experiments together.

the importance of hydrophobic interaction in this process. Nichols et al. have shown that protofibril growth, probably comparable to what is herein referred as protofibril fibrillization, occurs by two pathways [27]. They suggest that both protofibril elongation (i.e. monomer addition) and protofibril association occurs, the latter only in the presence of NaCl [27]. In addition, Stine et al. demonstrated, using EM, that oligomers coalesced into small aggregates in the presence of NaCl [22]. This could indicate a more significant component of protofibril association for Ab1–42Arc protofibril fibrillization, as this process was highly dependent on NaCl concentration. Ab concentration Ab1–42wt monomer oligomerization was found to be independent of the initial Ab concentration. This demonstrates that the early oligomerization phase is not a

nucleation-dependent mechanism, as the formation of nuclei (seeds) is expected to be a protein concentrationdependent process with a lag phase [28]. A similar theory has been suggested by Pallitto et al., where a nucleation mechanism was proposed for the late fibrillization phase, but not for the early oligomerization of monomers and dimers [29]. Interestingly, the maximum amount of protofibrils formed was increased with increasing Ab concentration, and a lag phase can be observed at 25 and 50 lm, before fibrillization into fibrils. We therefore argue that protofibrils could probably function as nuclei for fibril formation. The lag phase was also found to be dependent on temperature, indicating a thermodynamically unfavorable process, consistent with a nucleation mechanism. In agreement, Harper et al. suggested that dilution of nuclei can lead to fibrillar aggregates [30], which could explain the observed increased Ab1–42Arc protofibril fibrillogenesis at lower concentrations (i.e. higher dilution).


2625



Fig. 8. Cryo transmission electron microscopy (cryo-TEM) images of Ab1–42Arc and Ab42wt fibrils (A,C) and protofibrils (B,D). (A) Ab1– 42Arc fibrils incubated for 5 h under standard conditions [37 C, pH 7.8, 50 lM amyloid-b peptide (Ab)]. (B) Ab1–42Arc protofibrils incubated for 10 min under standard conditions. Inset: Ab1–42Arc protofibrils assembled into large dense clusters. (C) Ab1–42wt fibrils incubated for 48 h under standard conditions. (D) Ab1–42wt protofibrils incubated for 3.5 h under standard conditions. The fibrillar preparations do not contain protofibrils when centrifuged and analyzed with size-exclusion chromatography (SEC). The protofibrillar preparations contain protofibrils and small amounts of monomers when centrifuged and analyzed with SEC. Fibril samples analyzed with cryo-TEM were not centrifuged. Protofibril samples were centrifuged before cryo-TEM. Scale bar, 100 nm.

pH pH has significant consequences on many biochemical reactions owing to charge alterations of amino acid side chains. High and low pH are known to favour a-helical structure in the Ab peptide, in contrast to the pH interval 4–7, where the b-sheet structure is favoured [31]. In this study, the aggregation rates of Ab1–42wt and Ab1–42Arc were extremely rapid at pH 5.2, demonstrating the high aggregation rate at the isoelectric point (pH 5.3 for Ab1–42wt and pH 5.8 for Ab1– 42Arc) of the peptides, where they are close to uncharged. This is probably an effect caused by isoelectric precipitation. It is unclear if the aggregates formed at this pH are amyloid fibrils or amorphous aggregates. However, Ab1–42wt has been shown to form fibrils at acidic pH that are similar to fibrils formed at neutral pH [32].

2626

Raising the pH from pH 5.2 to pH 7.8 reduced the aggregation rate, both for Ab1–42wt and for Ab1– 42Arc. At this pH, the side chains of the histidines in positions 6, 13 and 14 are uncharged, whereas they are protonated (i.e. positively charged) at pH 5.2. Histidines are strongly involved in Ab aggregation. For example, substitution of His13 for Gln retards the aggregation process [33], nicotine binding to His reduces aggregation [34] and metals binding His residues are implicated in Ab aggregation [35]. Hence, the reduced aggregation rate at pH 7.8 compared with that at pH 5.2 is probably a result of charge alterations of His residues. Raising the pH even further, to pH 9.1, reduced the aggregation rate for Ab1–42wt monomer oligomerization, probably owing to the known stabilization of the a-helix structure at this high pH [31]. Interestingly, the amount of Ab1–42Arc protofibrils formed at pH 9.1 was essentially the same as at pH 7.8, demonstrating that the Arctic mutation overcomes the effect of high



pH on protofibril formation. Similar results have previously been reported for Ab with the Dutch mutation (E22Q), where the b-sheet conformation of the peptide persisted much more at higher pH compared with the wild-type peptide [36]. In summary, we demonstrate that the Arctic mutation significantly accelerates the aggregation process, and that the two distinct processes, monomer oligomerization and protofibril fibrillization, can be affected differently by changes in the micro-environment. Ab1–42wt monomer oligomerization was more energy dependent and involved hydrophobic interaction to a higher extent than protofibril fibrillization. Ab1– 42Arc protofibril fibrillization involved more hydrophobic interactions than wild-type protofibril fibrillization, and the process was rapid, even at high pH.

Experimental procedures Ab peptides Synthetic Ab1–42wt (lot H-J-1082, lot D-5003 and lot I-J1245) and Ab1–42Arc (lot G-0950RFW and lot G-A-1003) were purchased from PolyPeptide Laboratories GmbH (Wolfenbuttel, Germany). Recombinant Ab1–42wt purchased from rPeptide (Athens, GA, USA) was used for cryo-TEM. The lyophilized peptides were stored desiccated in glass vials. Siliconized tips and tubes were used for all peptide solutions (Sigma-Aldrich, St Louis, MO, USA). All peptides were dissolved to the desired concentration using their true peptide weight, as determined by the manufacturer.

Preparation of peptide Peptides were weighed on a Mettler Toledo balance AX26 Delta Range (Mettler Toledo, Stockholm, Sweden) and dissolved in 10 mm NaOH to a final concentration of 100 lm followed by vortexing for 2 min. This solution was divided into three tubes. While initiating the experiment for the first parameter, the two other NaOH ⁄ peptide stocks were kept in an ice bath (Ab1–42wt) or at )80 C (Ab1–42Arc). This means that for each experimental parameter, identical homogenous NaOH ⁄ peptide solutions were used. However, for the replicate experiments, new peptide was prepared fresh.

Experimental design Temperature Ice-cold 2 · phosphate-buffered saline (NaCl ⁄ Pi) (100 mm sodium phosphate, 200 mm NaCl, pH 7.4) was added to NaOH ⁄ peptide stock and the solution was vortexed for 1 min. The 50 lm peptide solution was divided into aliquots


and incubated at 37 C. The other stocks were diluted in 2 · NaCl ⁄ Pi and subsequently incubated at 25 C and 5 C.

Ab concentration Ice-cold 4 · NaCl ⁄ Pi was added to NaOH ⁄ peptide stock to a final Ab concentration of 75 lm, and vortexed for 1 min. The peptide solution was divided into aliquots and incubated at 37 C. The other stocks were diluted in 2 · NaCl ⁄ Pi and 1.33 · NaCl ⁄ Pi to final concentrations of 50 and 25 lm, respectively, and subsequently incubated at 37 C.

Ion strength Ice-cold 2 · NaCl ⁄ Pi with a high salt concentration (100 mm sodium phosphate, 1.0 m NaCl, pH 7.4) was added to NaOH ⁄ peptide stock to a final NaCl concentration of 0.5 m, and vortexed for 1 min. The 50 lm peptide solution was divided into aliquots and incubated at 37 C. The other stocks were diluted in 2 · NaCl ⁄ Pi to final NaCl concentrations of 0.1 m and 0 m, and subsequently incubated at 37 C.

pH Ice-cold 2 · acetate buffer (100 mm Na-acetate, 200 mm NaCl, pH 5.0) was added to NaOH ⁄ peptide stock and vortexed for 1 min. The 50-lm peptide solution was divided into aliquots and incubated at 37 C. The other stocks were diluted in 2 · NaCl ⁄ Pi (pH 7.4) and 2 · Tris Buffer (100 mm Tris ⁄ HCl, 200 mm NaCl, pH 9.0), and vortexed for 1 min. pH was measured with a Thermo Orion micro sodium electrode after addition of buffer. pH values were 5.16 (SD ¼ 0.03), 7.76 (SD ¼ 0.06) and 9.10 (SD ¼ 0.05).

SEC Before each SEC analysis, 1.8% Tween-20 was added to the sample to a final concentration of 0.6%, resulting in a final peptide concentration of 35 lm. The sample was centrifuged (5 min, 17 900 g, 16 C), to remove insoluble fibrillar material. Ten microlitres of supernatant was analyzed on a Merck Hitachi D-7000 HPLC LaChrom system with a diode array detector (VWR, Stockholm, Sweden). The SEC column used was a Superdex 75 PC3.2 ⁄ 30 column (GE Healthcare Bio-Sciences, Uppsala, Sweden). The samples were eluted at a flow rate of 0.08 mLÆmin)1, at ambient temperature, with NaCl ⁄ Pi containing Tween-20 (50 mm sodium phosphate, 0.15 m NaCl, pH 7.4, 0.6% Tween-20). All injected samples were subjected to wavelength scanning between 200 and 400 nm, and data were collected from 214 nm. Peak areas were integrated using Merck Hitachi model D-7000 Chromatography Data station software.


2627



fibrils, demonstrating that at this time point essentially all Ab had fibrillated.

Data analysis Data were fitted with a first-order exponential decay function. The goodness-of-fit varied between different experiments. The half lives derived from the fits are therefore an estimation of the aggregation rate and should be viewed as such. In the cases where no reasonable plateau could be calculated, this was set to 2 · 105 AU, which is a value close to zero in this assay. Half lives were derived from the curve fit, using graphpad prism v.3.0. All half lives are the mean half life from the individual experiments (unless otherwise stated). Statistical analysis [anova (n > 3) or unpaired t-test (n < 3)] was conducted using graphpad instat v.3.0.

Quantitative amino acid analysis Pellets formed from centrifugation at 17 900 g after incubating 50 lm Ab1–42Arc for 0–23 h in NaCl ⁄ Pi were dissolved in hexafluoroisopropanol (HFIP) and transferred to hydrolysis tubes. HFIP was evaporated at room temperature, and quantitative amino acid analysis was performed at the protein analysis center (PAC), Department of Medical Biochemistry and Biophysics, Karolinska Institutet, Stockholm, Sweden.

ThT fluorescence Pellets formed from centrifugation at 17 900 g after incubating 50 lm Ab1–42Arc in NaCl ⁄ Pi for 0–23 h were dissolved in 10 lm ThT, and ThT fluorescence was measured at kexcitation ¼ 440 nm and kemission ¼ 485 nm.

Cryo-TEM Peptide preparation Peptides were dissolved in 10 mm NaOH to a final concentration of 100 lm followed by vortexing for 2 min. The peptide solution was then neutralized with 2 · NaCl ⁄ Pi, followed by vortexing for 1 min.

Preparation of protofibrils and fibrils Ab1–42wt protofibrils were prepared by incubating 50 lm Ab for 3.5 h at 37 C, followed by centrifugation for 5 min at 17 900 g. At this time point, the sample typically consists of 83% protofibrils, according to SEC analysis. Ab1–42Arc protofibrils were prepared by incubating 50 lm Ab1–42Arc for 10 min at 37 C, followed by centrifugation for 5 min at 17 900 g. This typically results in a chromatogram of 95% protofibrils. Fibrils were prepared by incubating 50 lm Ab1–42wt and Ab1–42Arc for 48 h and 5 h at 37 C, respectively. This yielded a supernatant, after centrifugation at 17 900 g, which was essentially free of proto-

2628

Sample preparation for cryo-TEM and image recording The method consisted, in short, of the following (a more detailed description is available in Almgren et al. [37]). The samples were equilibrated at 25 C and 99% relative humidity within a climate chamber. A small drop ( 1 lL) of sample was deposited onto a copper grid covered with a perforated polymer film and with a thin carbon layer on both sides. Excess liquid was thereafter removed by means of blotting with a filter paper, leaving a thin film of the solution on the grid. Immediately after blotting, the sample was vitrified in liquid ethane, held just above its freezing point. Samples were kept below )165 C and protected against atmospheric conditions during both transfer to the transmission electron microscope and examination. The cryo-TEM investigations were performed with a Zeiss EM 902A transmission electron microscope (Carl Zeiss NTS, Oberkochen, Germany). The instrument operated at 80 kV and in zero loss bright-field mode. Digital images were recorded under low-dose conditions with a BioVision Pro-SM Slow Scan CCD camera (Proscan GmbH, Scheuring, Germany) and analysis software (Soft Imaging System, GmbH, Mu¨nster, Germany). In order to visualize as many details as possible, an underfocus of 1–2 lm was used to enhance the image contrast.

Acknowledgements This work was supported by grants from EU Consortiums Diadem and Apopis, Gun och Bertil Stohnes Stiftelse, Stiftelsen fo¨r Gamla tja¨narinnor, Alzheimerfonden, Hja¨rnfonden, the Swedish Research Council (project no. 2003–5546) and Bertil Ha˚llstens forskningsstiftelse.

References 1 Fratiglioni L, Launer LJ, Andersen K, Breteler MM, Copeland JR, Dartigues JF, Lobo A, Martinez-Lage J, Soininen H & Hofman A (2000) Incidence of dementia and major subtypes in Europe: a collaborative study of population-based cohorts. Neurologic Diseases in the Elderly Research Group. Neurology 54, S10–S15. 2 Walsh DM, Lomakin A, Benedek GB, Condron MM & Teplow DB (1997) Amyloid beta-protein fibrillogenesis. Detection of a protofibrillar intermediate. J Biol Chem 272, 22364–22372. 3 Harper JD, Wong SS, Lieber CM & Lansbury PT (1997) Observation of metastable Abeta amyloid protofibrils by atomic force microscopy. Chem Biol 4, 119–125.



4 Klein WL (2002) ADDLs & protofibrils – the missing links?. Neurobiol Aging 23, 231–235. 5 Rochet JC & Lansbury PT Jr (2000) Amyloid fibrillogenesis: themes and variations. Curr Opin Struct Biol 10, 60–68. 6 Walsh DM, Hartley DM, Kusumoto Y, Fezoui Y, Condron MM, Lomakin A, Benedek GB, Selkoe DJ & Teplow DB (1999) Amyloid beta-protein fibrillogenesis. Structure and biological activity of protofibrillar intermediates. J Biol Chem 274, 25945–25952. 7 Hartley DM, Walsh DM, Ye CP, Diehl T, Vasquez S, Vassilev PM, Teplow DB & Selkoe DJ (1999) Protofibrillar intermediates of amyloid beta-protein induce acute electrophysiological changes and progressive neurotoxicity in cortical neurons. J Neurosci 19, 8876–8884. 8 Nilsberth C, Westlind-Danielsson A, Eckman CB, Condron MM, Axelman K, Forsell C, Stenh C, Luthman J, Teplow DB, Younkin SG et al. (2001) The ‘Arctic’ APP mutation (E693G) causes Alzheimer’s disease by enhanced Abeta protofibril formation. Nat Neurosci 4, 887–893. 9 Klyubin I, Walsh DM, Cullen WK, Fadeeva JV, Anwyl R, Selkoe DJ & Rowan MJ (2004) Soluble Arctic amyloid beta protein inhibits hippocampal long-term potentiation in vivo. Eur J Neurosci 19, 2839–2846. 10 Selkoe DJ (2001) Alzheimer’s disease: genes, proteins, and therapy. Physiol Rev 81, 741–766. 11 Lashuel HA, Lai Z & Kelly JW (1998) Characterization of the transthyretin acid denaturation pathways by analytical ultracentrifugation: implications for wild-type, V30M, and L55P amyloid fibril formation. Biochemistry 37, 17851–17864. 12 Lashuel HA, Petre BM, Wall J, Simon M, Nowak RJ, Walz T & Lansbury PT Jr (2002) Alpha-synuclein, especially the Parkinson’s disease-associated mutants, forms pore-like annular and tubular protofibrils. J Mol Biol 322, 1089–1102. 13 Conway KA, Harper JD & Lansbury PT (1998) Accelerated in vitro fibril formation by a mutant alpha-synuclein linked to early-onset Parkinson disease. Nat Med 4, 1318–1320. 14 El-Agnaf OM, Nagala S, Patel BP & Austen BM (2001) Non-fibrillar oligomeric species of the amyloid ABri peptide, implicated in familial British dementia, are more potent at inducing apoptotic cell death than protofibrils or mature fibrils. J Mol Biol 310, 157– 168. 15 Poirier MA, Li H, Macosko J, Cai S, Amzel M & Ross CA (2002) Huntingtin spheroids and protofibrils as precursors in polyglutamine fibrilization. J Biol Chem 277, 41032–41037. 16 Whalen BM, Selkoe DJ & Hartley DM (2005) Small non-fibrillar assemblies of amyloid beta-protein bearing the Arctic mutation induce rapid neuritic degeneration. Neurobiol Dis 20, 254–266.


17 Lashuel HA, Hartley DM, Petre BM, Wall JS, Simon MN, Walz T & Lansbury PT Jr (2003) Mixtures of wild-type and a pathogenic (E22G) form of Abeta40 in vitro accumulate protofibrils, including amyloid pores. J Mol Biol 332, 795–808. 18 Van Nostrand WE, Melchor JP, Cho HS, Greenberg SM & Rebeck GW (2001) Pathogenic effects of D23N Iowa mutant amyloid beta-protein. J Biol Chem 276, 32860–32866. 19 Paivio A, Jarvet J, Graslund A, Lannfelt L & WestlindDanielsson A (2004) Unique physicochemical profile of beta-amyloid peptide variant abeta1–40E22G protofibrils: conceivable neuropathogen in Arctic mutant carriers. J Mol Biol 339, 145–159. 20 Harper JD, Wong SS, Lieber CM & Lansbury PT Jr (1999) Assembly of A beta amyloid protofibrils: an in vitro model for a possible early event in Alzheimer’s disease. Biochemistry 38, 8972–8980. 21 Snyder SW, Ladror US, Wade WS, Wang GT, Barrett LW, Matayoshi ED, Huffaker HJ, Krafft GA & Holzman TF (1994) Amyloid-beta aggregation: selective inhibition of aggregation in mixtures of amyloid with different chain lengths. Biophys J 67, 1216–1228. 22 Stine WB Jr, Dahlgren KN, Krafft GA & LaDu MJ (2003) In vitro characterization of conditions for amyloid-beta peptide oligomerization and fibrillogenesis. J Biol Chem 278, 11612–11622. 23 Naiki H & Nakakuki K (1996) First-order kinetic model of Alzheimer’s beta-amyloid fibril extension in vitro. Lab Invest 74, 374–383. 24 Kusumoto Y, Lomakin A, Teplow DB & Benedek GB (1998) Temperature dependence of amyloid beta-protein fibrillization. Proc Natl Acad Sci USA 95, 12277– 12282. 25 Gursky O & Aleshkov S (2000) Temperature-dependent beta-sheet formation in beta-amyloid Abeta (1–40) peptide in water: uncoupling beta-structure folding from aggregation. Biochim Biophys Acta 1476, 93–102. 26 Jarvet J, Damberg P, Danielsson J, Johansson I, Eriksson LE & Graslund A (2003) A left-handed 3(1) helical conformation in the Alzheimer Abeta (12–28) peptide. FEBS Lett 555, 371–374. 27 Nichols MR, Moss MA, Reed DK, Lin WL, Mukhopadhyay R, Hoh JH & Rosenberry TL (2002) Growth of beta-amyloid (1–40) protofibrils by monomer elongation and lateral association. Characterization of distinct products by light scattering and atomic force microscopy. Biochemistry 41, 6115–6127. 28 Jarrett JT & Lansbury PT Jr (1993) Seeding ‘one-dimensional crystallization’ of amyloid: a pathogenic mechanism in Alzheimer’s disease and scrapie? Cell 73, 1055– 1058. 29 Pallitto MM & Murphy RM (2001) A mathematical model of the kinetics of beta-amyloid fibril growth from the denatured state. Biophys J 81, 1805–1822.


2629



30 Harper JD & Lansbury PT Jr (1997) Models of amyloid seeding in Alzheimer’s disease and scrapie: mechanistic truths and physiological consequences of the timedependent solubility of amyloid proteins. Annu Rev Biochem 66, 385–407. 31 Barrow CJ, Yasuda A, Kenny PT & Zagorski MG (1992) Solution conformations and aggregational properties of synthetic amyloid beta-peptides of Alzheimer’s disease. Analysis of circular dichroism spectra. J Mol Biol 225, 1075–1093. 32 Su Y & Chang PT (2001) Acidic pH promotes the formation of toxic fibrils from beta-amyloid peptide. Brain Res 893, 287–291. 33 Kirkitadze MD, Condron MM & Teplow DB (2001) Identification and characterization of key kinetic intermediates in amyloid beta-protein fibrillogenesis. J Mol Biol 312, 1103–1119.

2630

34 Moore SA, Huckerby TN, Gibson GL, Fullwood NJ, Turnbull S, Tabner BJ, El-Agnaf OM & Allsop D (2004) Both the D-(+) and L-(-) enantiomers of nicotine inhibit Abeta aggregation and cytotoxicity. Biochemistry 43, 819–826. 35 Bush AI (2003) The metallobiology of Alzheimer’s disease. Trends Neurosci 26, 207–214. 36 Fraser PE, Nguyen JT, Inouye H, Surewicz WK, Selkoe DJ, Podlisny MB & Kirschner DA (1992) Fibril formation by primate, rodent, and Dutch-hemorrhagic analogues of Alzheimer amyloid beta-protein. Biochemistry 31, 10716–10723. 37 Almgren M, Edwards K & Karlsson G (2000) Cryo transmission electron microscopy of liposomes and related structures. Colloids Surfaces A: Physicochem Engineering Aspects 174, 3–21.
