Introduction

Many systemic and neurodegenerative disorders have been found to be associated with the misfolding and aggregation of specific proteins (Soto et al., 2006; Soto and Estrada, 2008). Alzheimer’s disease (AD) is one such degenerative disorder of the brain, accounting for about 50–70 % of the typical, late-onset cases of dementia (Goedert and Spillantini, 2006). AD is characterized by deterioration of cognitive function and behavioral changes and the wide-spread appearance of intraneuronal neurofibrillary tangles and the extracellular deposition of senile plaques (Hardy, 2006). These plaques are primarily composed of amyloid-beta peptide (Aβ). Genetic and neuropathological studies provide strong evidence for the central role of Aβ peptide in the pathogenesis of AD (Hutton et al., 1998; Annaert and De Strooper, 2002; Taylor et al., 2002; Selkoe, 2004). Aβ is a 39–42 residue peptide derived from the amyloid precursor protein (APP) that is normally cleaved by the proteases β- and γ-secretases (Findeis, 2007). Aβ undergoes aggregation spontaneously and assembles into 6–10 nm diameter fibrils with a cross beta-sheet structure (Harper et al., 1999; Goldsbury et al., 2000). Numerous observations indicate that the aggregation of Aβ into fibrils has neurotoxic effects in cell culture and in vivo (Yankner, 1996; Hartley et al., 1999; Pike et al., 1991, 1993; Lorenzo and Yankner, 1994). However, more recent studies demonstrated that prefibrillar intermediates are also capable of initiating pathogenic events (Lambert et al., 1998; Koo et al., 1999; Hartley et al., 1999; Lansbury, 1999; Ward et al., 2000; Dahlgren et al., 2002; Kayed et al., 2003; Glabe, 2006; Walsh and Selkoe, 2007; Hoozemans et al., 2006). This suggests that targeting a specific step involved in Aβ aggregation might inhibit Aβ neurotoxicity. Considerable research efforts have focused on developing treatments that block the toxicity of Aβ. These include the use of protease inhibitors to prevent Aβ production (Beher and Graham, 2005; Vassar, 2005; Wolfe, 2007), immunization with Aβ antibodies (Solomon, 2004; Petrushina et al., 2007), and use of small molecules and peptides to prevent Aβ aggregation (Weggen et al., 2007; Austen et al., 2008; Rivière et al. 2008; Ladiwala et al., 2011a, b). Previous studies showed that the central hydrophobic region of Aβ (residues 16–20 of Aβ, KLVFF) is crucial for Aβ polymerization (Tjernberg et al., 1996). Based on this study, several groups have designed peptides that modulate Aβ fibrillogenesis and toxicity (Soto et al., 1998; Findeis et al. 1999; Gordon et al., 2001; Lowe et al., 2001). d-amino acid peptides have also been identified as inhibitors of amyloid formation and cytotoxicity of the Aβ peptide (Tjernberg et al., 1997; Blanchard et al., 1997; Chalifour et al., 2003; Wiesehan et al., 2003). The major disadvantage with the use of l-amino acid peptides as inhibitors with respect to d-amino acid peptide is their inactivation by proteolysis (Milton et al., 1992). Taken orally, d-peptides can be absorbed systemically whereas l-peptides must be injected to avoid digestion (Pappenheimer et al., 1994, 1997). In this work, we investigated the potential effectiveness of d-amino acid peptides containing Aβ self-recognition motif (amino acid residues, 16–20) on Aβ aggregation and toxicity. We used an in vitro model of Aβ to show the effects of these peptides on Aβ fibrils as well as oligomers formation while in vivo studies were conducted in transgenic Caenorhabditis elegans model that express the human Aβ42 minigene driven by the unc-54 promoter (strain CL2006), Aβ42 is solely expressed within the body wall muscles, resulting in their paralysis (Wu and Luo, 2005; Link, 2006).

Results

Effects of d-peptides on aggregation/turbidity of Aβ

The aggregation of Aβ42 in the presence of d-peptides was monitored by turbidity measurements as shown in Fig. 1. The absorbance of Aβ42 solution measured at a wavelength of 405 nm was used to estimate the amount of aggregation. Both kklvffa and kklvffarrrra at 1:1 and 1:5 molar ratios showed an increase in turbidity of solution. However, the presence of pgklvya under similar conditions led to a moderate decrease in the extent of Aβ42 aggregation as compared to peptide incubated alone.

Fig. 1
figure 1

Aggregation of Aβ42 in the absence or in the presence of kklvffarrrra, kklvffa, and pgklvya as measured by observing turbidity at 405 nm. Data are mean ± SEM

Inhibition of fibril formation

The fibril formation of Aβ was measured by ThT binding on aggregates after incubation of Aβ solution in the presence of the d-peptides as shown in Fig. 2. Incubation of Aβ42 with kklvffarrrra and kklvffa resulted in low ThT fluorescence levels suggesting the lack of fibril formation. Furthermore, pgklvya under similar conditions did not have any significant effect on Aβ42 fibrillization. d-peptides in the absence of Aβ did not show an increase in ThT fluorescence.

Fig. 2
figure 2

Effect of the d-peptides on Aβ fibrillogenesis. Samples of Aβ42 incubated for 3 days in the absence or in the presence of d-peptides. The amyloid fibril formation was then measured by the Thioflavin T binding assay. Data are mean ± SEM

ThT-bound Aβ fibril was also visualized by fluorescence microscopy (Ban et al., 2004). Aβ42 incubated in the buffer alone shows fibrils of few micrometers (Fig. 3a). Co-incubation with kklvffarrrra and kklvffa led to the formation of a small number of fibrils (Fig. 3b, d). In contrast, Aβ42 aggregates with pgklvya showed mixed morphology with both fibrils and spherical aggregates (Fig. 3c).

Fig. 3
figure 3

Fluorescence microscopy images. ThT stained samples of Aβ42 peptide incubated in the absence a or in the presence of kklvffarrrra (b), pgklvya (c), and kklvffa (d). The scale bar represents 20 μm

ANS fluorescence

To probe the nature of the interaction between Aβ42 and the d-peptides, bis-ANS was used. bis-ANS is a fluorescent dye which shows an increase in fluorescence intensity and blue-shifting of the emission maximum upon binding to the solvent-exposed hydrophobic surfaces of Aβ aggregates (Stryer 1965; Turner and Brand 1968). As shown in Fig. 4, Aβ in its monomeric form had no effect on ANS fluorescence (Kremer et al., 2000). Incubating Aβ42 samples for 3 days showed an increase in ANS fluorescence intensity and a blue shift of the emission maxima to ~489 nm. The ANS fluorescence intensity of Aβ42 increased remarkably when co-incubated with kklvffa and kklvffarrrra while only a small increase was detected when incubated with pgklvya. This means that kklvffarrrra and kklvffa induce exposure of hydrophobic sites in Aβ.

Fig. 4
figure 4

Effect of Aβ42 aggregations on bis-ANS fluorescence in the presence or in the absence of d-peptides

Circular dichroism absorption spectroscopy studies

Circular dichroism spectroscopy (CD) was used to determine the effects of d-peptides on the secondary structure of amyloid-β peptide (Aβ). Aβ peptide on aging undergoes the conformational transition from the unordered structures (random coil) to the β-sheet conformation and is associated with neurotoxic activity in vitro (Simmons et al., 1994; Gordon et al., 2001). The CD spectrum of monomeric Aβ42 peptide was characterized by a strong negative band in the 195–200 nm region, indicative of a disordered or random coil conformation (Fig. 5). After incubating for 3 days, Aβ42 peptide generates β-sheet conformation with a minimum around 218 nm. On the other hand, after 3 days Aβ42 in the presence of kklvffa, kklvffarrrra, and pgklvya remained in random coil conformation with a small reduction in the intensity of the band at 198 nm.

Fig. 5
figure 5

Far-UV CD spectra of Aβ in the absence or in the presence of d-peptides. The CD measurements were carried out at 25 °C, Aβ42 alone at 0 h (black), Aβ42 at 72 h (green), Aβ42/kklvffa (blue), Aβ42/kklvffarrrra (pink), Aβ42/pgklvya (brown) (Color figure online)

Effects of d-peptides on soluble aggregates of Aβ (oligomers)

A growing number of reports revealed that the most toxic agent of Aβ is nonfibrillar and oligomeric in nature (Lambert et al., 1998; Walsh et al., 2002; Gong et al., 2003). Therefore, compounds which are able to inhibit their formation helps in reducing Aβ toxicity. Oligomers of Aβ42 were characterized in vitro by silver stain. As shown in Fig. 6, lane 1 represents Aβ42 initial aggregation species. The molecular masses of ~16, ~22 kDa, correspond to oligomers of Aβ42. On the other hand, incubation of Aβ42 with kklvffa and pgklvya inhibits the formation of oligomers, while Aβ42 solution incubated with kklvffarrrra shows bands around ~50 and ~98 kDa representing large oligomeric species.

Fig. 6
figure 6

Oligomerization of Aβ in the absence or in the presence of d-peptides at 37 °C for 48 h. Samples incubated at 37 °C for 48 h and after centrifugation 10 μl of the supernatant were separated by SDS-PAGE on a 12 % Nupage Bis–Tris gel and visualized with silver stain. Aβ42 oligomer alone (lane 1), Aβ42/kklvffa (lane 2), Aβ42/pgklvya (lane 3), Aβ42/kklvffarrrra (lane 4)

Survival of transgenic C. elegans fed with d-peptides

Transgenic C. elegans expressing human Aβ42 were employed to study the effects of d-peptides on Aβ toxicity in vivo (Link, 1995; Link et al., 2001). Treatment of the worms with 124 μg/ml kklvffa and 124 μg/ml pgklvya resulted in prolonged survival compared to the control animals (Fig. 7a, b). On the other hand, treatment with 123 μg/ml kklvffarrrra did not have significant effect on survival in nematodes.

Fig. 7
figure 7

Survival of the transgenic C. elegans strain CL2006 in the presence of d-peptides. (** < 0.01, * < 0.05)

Discussion

There is now substantial evidence from genetics and biochemical studies that conversion of a soluble Aβ protein into pre-fibrillar, diffusible assemblies and further fiber formation are a key event in the pathogenesis of AD. Therefore, compounds that interfere with this process may be the viable strategy for therapeutic intervention in AD. Earlier studies indicate that peptide’s chirality is not critical for peptide–peptide interactions (Pritsker et al., 1998). l- and d-peptides derived from type IV collagen bind the α3β1 integrin (Milton RC Milton SC Kent 1992). d-peptides derived from chemokines act as antagonists to CXCR4 receptor, resulting in inhibition of CXCR4-dependent HIV-1 (Zhou et al., 2002). Similarly, both d- and l-form of peptides compete for the same binding site in DnaJ chaperone system (Feifel et al., 1998; Bischofberger et al., 2003). In addition, all-d and all-l conformers of the functional element of alphaA-crystallin did not show marked differences in their chaperone-like activity (Bhattacharyya and Sharma, 2001). Furthermore, peptides in d-forms are known to be more protease resistant and less immunogenic than the respective l-enantiomers, giving them additional therapeutics value. Antimicrobial peptides in d-form showed significant improvements in the therapeutic indices when compared with the l-enantiomers (Chen et al., 2006). In another study, the impairment of learning and memory performance of apolipoprotein E gene knockout mice was attenuated with d-amino acid peptide treatment (Brenneman et al., 2004). d-peptides have also been identified as inhibitors of amyloid formation and cytotoxicity of the Aβ peptide (Blanchard et al., 1997; Wiesehan et al., 2003). d-peptides containing self-recognition motif of the Aβ are more effective in inhibiting fibrillogenesis and reducing the toxicity of Aβ in neurons (Chalifour et al., 2003).

In this study, the aim of the work is to design short synthetic peptides consisting of d-amino acids to investigate their influence on Aβ aggregation and Aβ induced pathological behaviors in transgenic C. elegans model. These peptides were designed from the region of Aβ (amino acid residues, 16–20) that is responsible for its self-association and aggregation (Tjernberg et al., 1996; El-Agnaf et al., 1998). Aggregation processes are affected by hydrophobicity, β-sheet propensity, and charge. Single amino acid substitutions affect the amyloid potential of several peptides and proteins (Otzen et al., 2000; Villegas et al., 2000; Giasson et al., 2001; Chiti et al., 2002, 2003; Tjernberg et al., 2002). Addition of proline, lysine, and arginine in the sequence is to “cap” the aggregation prone region of amyloid-β peptide and impair its aggregating tendency. These residues owing to their very low hydrophobicity, charge, and a β-sheet propensity disfavor protein aggregation (Schobert and Tschessche 1978; Srinivas and Balasubramanian, 1995; Samuel et al., 1997; Shiraki et al., 2002; Reddy et al., 2005).

Results from the ThT and turbidity assays suggested that in addition to the β-sheet aggregate, another type of aggregate is formed in the presence of d-peptides. Peptides kklvffa and kklvffarrrra inhibited Aβ fibril aggregates, but increased Aβ over all aggregations. This means that these peptides actively reduce amyloid fibrils and form insoluble amorphous aggregates. On the other hand, peptide pgklvya had no effect on Aβ fibrillogenesis as observed by ThT. However, in fluorescence images, the presence of spherical aggregates indicates that pgklvya may have triggered another mechanism of assembly of Aβ.

CD spectrum analysis suggested that d-peptides inhibited Aβ to adopt β-sheet conformation. This means that aggregates formed by Aβ in the presence of d-peptides had less ß-sheet content. Partially folded structures are generally the key precursor to protein aggregation (Kelly 1996; Fink 1998). Therefore, the nature of the intermediate state is essential for the understanding of the aggregation pathway. These folded structures are characterized by the presence of solvent-exposed hydrophobic clusters as indicated by ANS binding. Data from the result showed that peptides kklvffarrrra and kklvffa cause conformational change that leads to increased exposure of hydrophobic regions as observed by enhanced ANS binding. Hydrophobic sites are believed to play a major role in protein–protein interaction. This suggests that peptides bind to β-amyloid peptide, leading to additional interaction between the exposed hydrophobic sites, which accelerates assembly of Aβ peptide and forms amorphous aggregates. Several reports have demonstrated that compounds that interfere with Aβ aggregation can exert beneficial effects in vivo, in the mouse model of AD (Zou et al., 2003; Stackman et al., 2003; Yang et al., 2005). In this study, the pharmacological effects of d-amino acid peptides on Aβ-induced toxicity were tested on transgenic C. elegans.

Caenorhabditis elegans is an ideal model organism for useful analysis of the age-associated neurodegeneration because of its available genetic information as well the simple structure of its nervous system. Results from the assays indicate that among the three d-peptides tested, kklvffa and pgklvya prolonged survival in the transgenic Caenorhabditis elegans. The protective effect of these two d-peptides is associated with the inhibition of Aβ oligomeric species. Several studies of Aβ neurotoxicity have focused on the effects of insoluble amyloid fibrils; recent work suggests that accumulation of soluble pre-fibrillar aggregate of Aβ is more toxic to cultured neurons. In addition, a behavioral and morphological alteration found in the transgenic mice model of AD indicates that soluble oligomeric Aβ species can mediate neuronal injury in these animals (Hsia et al., 1999; Mucke et al., 2000). Based on the results, oligomeric species of Aβ is inhibited by pgklvya and kklvffa while kklvffarrrra inhibited (~20 kDa) oligomers but favors the formation of higher order of Aβ oligomers (~50 kDa) reported to be abundant in the AD brain (Gong et al., 2003). This suggests that kklvffarrrra although inhibited Aβ fibrils, did not improve survival in worms. Recently, it is reported that HSP-16.2 (small chaperone protein), which shows significant protection against Aβ toxicity in vivo, is associated with reduction of Aβ oligomeric protein in vitro (Fonte et al., 2008; Wu et al., 2006). In another study, EGb-71 suppresses Aβ toxicity by modulating Aβ oligomeric species (Wu et al., 2006).

Conclusion

In summary, using several methods, the effect of designed d-peptides on the early and later stage of Aβ aggregation was tested. The results suggested that kklvffarrrra and kklvffa are very effective at inhibiting fibrillogenesis of Aβ. Transgenic C. elegans model was used to evaluate the pharmacological effect of d-peptides on Aβ-initiated toxicity. Among the three peptides tested, only pgklvya and kklvffa are capable in improving survival in the transgenic C. elegans. The activity of these peptides correlates with their ability to inhibit toxic smaller oligomers of Aβ generated in the early phase of amyloidogenesis. These suggest that these d-peptides should be considered as potential therapeutics against Alzheimer’s disease.

Materials and methods

Source of chemicals

Dicyclohexylcarbodiimide (DCC) and Hydroxybenzotriazole (HOBt) were obtained from Sigma-Aldrich Chemical Company (St., Louis, Missouri, USA). Pentafluorophenol was obtained from Spectrochem Pvt. Ltd. (Mumbai, India). Pyridine, piperidine, phenol, and acetic anhydride were purchased from SRL (Mumbai, India). Sodium chloride, sodium dihydrogen phosphate, and disodium hydrogen phosphate were purchased from S.D. fine chemicals, India. DMAP (4-dimethylamino pyridine) and ninhydrin were obtained from Merck Chemicals (Mumbai, India). The reagents trifluoroacetic acid, ethane dithiol, and thioanisole were obtained from Aldrich (St., Louis, Missouri, USA) and were used as obtained. The solvent’s N, N-dimethylformamide (DMF), dichloromethane, methanol, ethanol, tetrahydrofuran, ethyl acetate, glycerol, hexane, and diethyl ether were all of A. R grade (Merck chemicals, India, Bombay). Wang resin and 9-fluorenylmethoxycarbonyl amino acids (Fmoc-d and Fmoc-l amino acids) for solid phase peptide synthesis were obtained from Novabiochem (Mumbai, India).

Synthesis of amyloid-β peptide and d-amino acid peptides by solid phase peptide synthesis

Amyloid-beta (Aβ) peptide and d-amino acids peptides were synthesized manually by Fmoc (9-fluorenylmethoxy carbonyl) protocols (Fields and Noble, 1990; Chan and White, 2000). The Wang resin was used as a solid support to build the peptide. Pentafluorophenyl active esters/HOBt esters/anhydrides of N-protected Fmoc amino acids (Fmoc-AA) were used as building units. Completion of coupling was monitored by Kaiser’s test (Kaiser et al., 1970) after each coupling. Piperidine solution (20 %) in DMF was used for deprotection of N-terminal Fmoc group followed by next coupling step. The peptide was cleaved from the resin support using 80–90 % trifluoroacetic acid (TFA) and precipitated in cold ether. Peptide was purified by reverse-phase HPLC on C18 chromatography columns using acetonitrile–water–trifluoroacetic acid gradients. Purity of the peptide was evaluated by Maldi-MS.

Aβ sample preparation

To get a monomeric Aβ42, the lyophilized peptide was solubilized and disaggregated using TFA, TFE/HFIP procedures (Chen and Wetzel, 2001). Then the peptide solution was aliquoted out, and HFIP was removed by evaporation and the peptide film was stored at −80 °C until required. Immediately, prior to use, the peptide was dissolved in a phosphate buffer, pH 7.4, to a required concentration.

Turbidity measurements

A simple way to measure the presence of aggregates in a solution is to determine the turbidity by measuring absorbance. By this method, all aggregations of Aβ were measured in the presence of d-peptides. An increase in turbidity would arise from the formation of large aggregates. Aβ42 at the concentration of 50 μM was incubated in the presence and in the absence of d-peptides for 3 days at 37 °C. At the end of incubation each sample was assayed by turbidity at 405 nm wavelength to estimate the amount of aggregations. All the samples containing d-peptides, background signals were determined by measuring the wavelength at 405 nm of samples containing only d-peptides. The resulting values were then used to correct each of the turbidity readings corresponding to assembly reactions.

Thioflavin T assay

Thioflavin T (ThT) is a benzothiazol salt used to visualize amyloid plaques. Free ThT has excitation and emission maxima at ~350 and ~450 nm. However, upon binding to fibrils, the excitation and emission wavelength change to ~450 and ~485 nm, respectively (Naiki et al., 1989, 1990; LeVine, 1999). A 100 μM solution of ThT was prepared and filtered through a 0.2 μM filter. Aβ42 at the concentration of 50 μM was incubated at 37 °C for 3 days either alone or in the presence of d-peptides. A 10 μl sample from the incubated mixture was added to solutions containing 10 μM ThT in phosphate buffer, pH 7.4, and the ThT fluorescence was measured at λ ex = 450 nm and λ ex = 485 nm using a FluoroLog-3 spectrofluorometer (HORIBA Jobin–Yvon, Edison, New Jersey, US).

Fluorescence microscopy

A fluorescence microscope is a light microscope used to study specimens, which can be made to fluoresce. Aβ42 at the concentration of 50 μM was incubated at 37 °C for 3 days either alone or in the presence of d-peptides. 10 μl samples from the incubated mixture were mixed with 5 μl of 50 μM ThT in phosphate buffer, pH 7.4, and the mixture was placed on a glass slide. The ThT-bound amyloid fibrils were observed with a Lieca DMRXA2 upright microscope equipped with a Leica I3 filter and a Leica DC500 camera (Leica Microsystems, Wetzlar, Germany). Images were collected using either a 10× or 60× oil-immersion lens.

ANS fluorescence

4, 4′-Bis (1-anilinonaphthalene 8-sulfonate) (bis-ANS) is frequently used to demonstrate the presence of partially folded conformations of proteins. Bis-ANS binds to solvent-exposed hydrophobic clusters, resulting in a considerable increase in the ANS fluorescence intensity and in a pronounced blue shift of the fluorescence emission maximum. We measured the change of ANS fluorescence to monitor the gain of structure in amyloid-beta peptide (Aβ) in terms of solvent-exposed hydrophobic clusters upon binding with the d-peptides. Aβ42 at the concentration of 50 μM was incubated for 3 days at 37 °C either alone or in the presence of d-peptides. A 10 μl sample from the incubated mixture was added to a solution containing 5 μM ANS in phosphate buffer, pH 7.4, and the fluorescence was measured using a FluoroLog-3 spectrofluorometer (HORIBA Jobin–Yvon, Edison, New Jersey, US). Emission spectrum was recorded from 400 to 600 nm with excitation at 370 nm.

Circular dichroism measurements

The effects of d-peptides on Aβ42 conformation were determined by CD. Aβ42 at a concentration of 50 μM was incubated either alone or in the presence of d-peptides at a molar ratio of 1:1. A spectrum was acquired on a Jasco Model J-715 spectropolarimeter. Dry purified nitrogen was employed to keep the instrument oxygen free during the experiments. The spectra were obtained using cylindrical fused quartz cells of 1 mm path length. A complete base line was recorded for each measurement using the same cell in which the sample solution had been replaced with pure solvent. The spectropolarimeter was calibrated using ammonium- d10-camphorsulfonate as described by the instrument manufacturers.

CD measurements settings

A spectrum was recorded with 1 nm bandwidth and 0.2 nm step resolution at a scan speed of 100 nm/min over the range of 190–260 nm. The spectrum of the peptides was recorded in phosphate buffer pH 7.4. The circular dichroism data represents average values from at least five to ten separate recordings. The resulting spectrum was smoothened. All spectra were corrected by subtraction of any contribution from buffer and d-peptides.

Preparation of Aβ soluble aggregates (oligomers)

42 stock solutions were obtained by dissolving the lyophilized peptide in anhydrous dimethyl sulfoxide (DMSO) followed by sonication for 30 s. The oligomerization reaction was initiated by diluting the stock solution in the phosphate buffer, pH 7.4 to a final concentration of 50 μM. The reactions were incubated at 37 °C for 2 days in the absence or in the presence of d-peptides. After incubation, samples were spun at 14,000×g (4 °C for 10 min) and the supernatant samples (10 μl) were electrophoresed at 100 V on a NuPAGE 4–12 % Bis–Tris SDS gel. The presence of Aβ soluble aggregate was confirmed by silver staining, according to manufacturer’s instructions (Fast silver, G-Biosciences, US).

Nematode strain

The muscle-specific Aβ-expressing strain was obtained from the Caenorhabditis Genetics Center. Large scale synchronized populations were prepared by alkaline hypochlorite egg purification and propagation of staged animals on Petri plates containing nematode growth media supplemented with 2 % peptone.

Caenorhabditis elegans survival assay

Age synchronized populations of CL2006 worms were grown on NGM IPTG Amp plates containing either a vehicle or d-peptides. To prevent progeny-production pos-1 bacterial strain was used as the food source. The bacteria colony was grown at 37 °C in LB with 100 μg/ml ampicillin, and then seeded on NGM IPTG Amp plates. All d-peptides were added directly to the Pos-1, and animals were exposed to d-peptides starting from the egg.