Although previous studies performed by our group have demonstrated the key role of the C-terminal domain of CETP during the process of lipid transfer when holding an α-helix conformation (residues E₄₆₅–S₄₇₆) [8,13], it has not been determined yet whether this domain presents the ability to undergo conformational changes. Therefore, in order to investigate if this domain might undergo conformational changes by modifying the physicochemical properties of the surrounding media, two peptides were synthesized: the native C-terminal peptide called helix-X, and a second one containing the mutation D₄₇₀N and named helix-Z (Figure 1).

Helix-X and helix-Z were incubated in water in a pH range of 3.8–9.5 and studied by circular dichroism (CD) spectroscopy, in parallel. The secondary structure of helix-X was consistent with the presence of a disordered state over the whole range of pH values tested (Figure 2A). In contrast, CD spectra of helix-Z showed a pH specific response with minimum shift signals from 219 nm to 201 nm, which is associated with an increase in ellipticity indicative of the formation of β-type secondary structures (Figure 2B). This spectrum was only found in a pH range of 7–8, reaching the highest value at pH 7.2 (Figure 2B insert).

Moreover, through the use of thioflavin-T (ThT) fluorescence assays, an increase in the emission spectrum of helix-Z with a maximum signal near 482 nm was observed (Figure 2C). Also, by testing with a Congo red-shift assay, the characteristic birefringence change observed in the presence of β-structures with a maximum signal close to 540 nm was determined (Figure 2D). Therefore, these assays confirm the possibility that under specific conditions the C-terminal domain of CETP can acquire a secondary structure consistent with the formation of a β-structure.

During the evaluation of the stability of helix-Z through temperature-induced unfolding CD assays (4–90 °C), only temperatures close to 65 °C were found to induce significant loss in its β-type secondary structure, suggesting high stability. The presence of an isodichroic point near 211 nm in the obtained spectra is associated with the presence of only two conformational states (Figure 3). These experiments, together with the fact that helix-Z keeps its characteristic CD profile consistent with a β-strand under a range of high ionic strength solutions (up to 2.4 M NaCl, data not shown), show that the acquired β-structure is not determined by long-range electrostatic interactions, but rather by stronger inter- and intra-chain interactions such as hydrogen bonds.

Another set of experiments, where both helix-X and helix-Z were incubated in a hydrophobic environment composed of sodium dodecyl sulfate (SDS) micelles, shows that both peptides acquire and maintain an α-helical conformation (Figure 4A,B). This transition was monitored in parallel following fluorescence coupled to ThT (Figure 4C) and the change in absorbance of peptide bonds at 205 nm (Figure 4D). Our results suggest that the interaction within a hydrophilic/hydrophobic interface may be the key parameter to maintain both peptides in an α-helix conformation; more importantly, in the case of helix-Z, this is a key feature to avoid the formation of a β-structure.

In order to extend our structural analysis and considering that an increase in β-signal is not a direct indication of the formation of amyloid fibrils, peptide samples were analyzed by transmission electron microscopy (TEM). Throughout the pH-range evaluated (4–9), helix-X samples did not form any kind of structured material, which correlates well with previous results in which helix-X remains in a non-ordered state (Figure 5A). In contrast, when helix-Z samples were analyzed, even from the first sample without incubation (time 0), small oligomers were clearly identified (Figure 5B,C). It was only after 1 h of incubation that protofilament formation was observed (Figure 5D).

Incubation of helix-Z for 6 h with stirring (200 rpm) was the condition that identified a mixture of oligomeric structures, protofilaments, and mature fibrils in some areas (Figure 5E). Samples analyzed after an incubation time of 72 and 120 h at 37 °C clearly showed mature fibrils with high morphological similarities to β-amyloid samples (Figure 5F,G). At this stage, it is interesting to mention that although the formation of fibrils from pH 4–9 was evaluated, aggregates and fibril formation were only observed at pH 7–8, which correlates well with the presence of β-structures at this pH range when monitored by spectrophotometric techniques.

Taking into account previous studies from several laboratories where microglia and macrophages were employed in the characterization of the cytotoxic effects of the β-amyloid peptide [15–17], when microglial cells were exposed to a gradual increase in concentrations of both peptides, only helix-Z was able to significantly reduce cell viability as measured by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) reduction assay (Figure 6). It is interesting to observe that independently of the incubation time used for both peptides in solution (0, 6, 120 h) the lowest concentration employed for helix-Z (1.7 μM) shows an important reduction in cell viability that is maintained up to 56 μM (Figure 6A–C). At this stage it is important to mention that 0 time actually corresponds to the time employed to place both peptides, first in solution and second into the cell culture plate. This procedure takes approximately 15 min, enough time to apparently produce peptide oligomers as shown in Figure 5B–C. Because cell viability is maintained close to 50% independently of the concentration of helix-Z or the time of the incubation needed for the formation of amyloid structures, our experiments indicate that helix-Z in the form of oligomers seem to be the highest toxic form for this peptide. Whether or not mature fibrils are also responsible for cell toxicity after the incubation of helix-Z for 120 h, will have to be elucidated in experiments where under these conditions the presence of small amounts of oligomers can be ruled out. At this stage it is interesting to mention that while helix-Z oligomers formed at pH 7.2 are cytotoxic, amorphous aggregates of the β-amyloid peptide obtained at the same pH do not present the ability to damage microglial cells in culture (Figure 1 supplementary data). Changes found in cell morphology are possibly related to the presence of a cellular stress condition as described by our group when studying the β-amyloid peptide [17] (Figure 6D). It is noteworthy that the MTT assay is also an indicator of oxidative stress and changes in the vesicular trafficking of cells [18,19].

After an incubation period of 120 h at 37 °C (condition at which mature fibrils are obtained with helix-Z) microglial cells were treated for 20 h with increasing concentrations of peptides. The experiment was focused on assessing the expression levels of proteins such as β-adaptin, eps 15, and clathrin assembly lymphoid myeloid leukemia (CALM). Cells exposed to helix-Z showed differential expression of some of these endocytic proteins, in contrast to experiments performed with helix-X in which no changes were found. While the expression of proteins such as eps 15 was held constant, β-adaptin protein expression decreased with respect to exposure to increasing concentrations of helix-Z (Figure 7A). In contrast, CALM protein expression increased with respect to the concentration of helix-Z (Figure 7B). These results, as shown previously by us, correlate well with the β-amyloid peptide (Aβ_25–35) experiments that were employed under the same conditions as used here with helix-Z (Figure 1 supplementary data).

All cell culture reagents were purchased from Gibco-Invitrogen (Carlsbard, CA, USA), while tissue culture dishes and other plasticware were obtained from Nulgene Nunc (Rochester, NY, USA). Salts and buffers were purchased from Baker. Tert-butyl hydroperoxide (TBH), ThT, Congo-red, SDS and MTT were obtained from Sigma (St. Louis, MO, USA). Antibodies used in CALM, β-adaptin, and eps 15 protein detection as well as goat secondary antibodies were from Santa Cruz Biotechnoloy Inc.

Based on the function of the C-terminal of CETP, two peptides were synthesized: one named helix-X that corresponds to the native peptide, and the second one named helix-Z containing the mutation D₄₇₀N. Lyophilized peptides were dissolved in ammonium carbonate buffer (pH 9.5) to a concentration of 1 mg/mL. From this solution a further 1:5 dilution was carried out. To evaluate the structure at pH 3.8 and 4.8, a sodium acetate buffer was used; pH 6.3 and 7.2, a sodium phosphate buffer was used, and at pH 8.6 and 9.5, an ammonium carbonate buffer was employed.

Fragment Aβ_25–35 (β-amyloid) was employed as a reference molecule for the cytotoxicity assays. Experiments performed with β-amyloid were carried out at two pH conditions using a phosphate buffer (pH 7.2) and ultrapure water (pH 5.5), both with the same peptide concentration (1.5 mg/mL). All buffers were used at 50 mM, and solutions were filtered through 0.22 μm membrane filters before use.

Purity of peptides was greater than 98% (GenScript Corp., Piscataway, NJ, USA). Their identity and purity were confirmed by mass spectrometry and HPLC analysis (data not shown). Protein concentration was determined by peptide bond absorption at 205 nm.

Far-UV CD spectra were recorded on an Aviv 62DS spectropolarimeter in a 0.1 cm quartz cell, using an average time of 2.5 s and a step size of 0.5 nm. CD results are reported as mean molar ellipticity (Θ, deg cm² dmol⁻¹) considering the baseline correction. For the pH studies, CD spectra were obtained from pH 3.8 to −9.5. Peptide unfolding induced by temperature was monitored by CD measurements over a temperature range of 4 to 90 °C.

Congo red assays were performed based in the protocol designed by Klunk et al. [35]. Employing 10.6 μM Congo red and 180 μg/mL of peptide, the absorbance was measured every 2 nm using a Perkin Elmer UV/Vis Lambda 2S spectrophotometer scanning from 360 to −700 nm. Peptide spectra were corrected by subtracting the corresponding buffer spectra obtained under identical conditions. Additionally, the presence of β-structures was characterized with the ThT assay. ThT fluorescence emission spectra were registered at 37 °C from 470 to 530 nm with an excitation wavelength of 450 nm. A scan velocity of 73 nm/min using an Olis DM45 spectrofluorimeter was used. The concentrations of ThT and peptides were 10 μM and 36 μM, respectively.

EOC cells (mouse microglia, American Type Culture Collection) were grown in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum, and 20% mouse bone marrow, producing colony stimulating factor-1 (LADMAC) conditioned media. Penicillin (50 U/mL) and streptomycin (50 μg/mL) were added to the media. After the initial culture, all experiments were performed using Opti-MEM medium without phenol red and reduced fetal serum.

Peptide cytotoxicity was assessed by MTT reduction assays in EOC cells exposed to helix-X, helix-Z, and the β-amyloid peptide (Aβ_25–35). Cells were seeded into 96-well plates at a density of 1 × 10⁴ cells/well and allowed to grow to 80% confluence. Next, culture medium was replaced with Opti-MEM medium. After 2 h under this condition, cells were treated with a series of gradually increasing peptide concentrations for 20 h. After this time, 30 μL of an MTT stock solution in Opti-MEM medium was added to the cultures in order to obtain a final concentration of 0.5 mg/mL. Formazan crystals that formed after 4 h of incubation were further dissolved by the addition of cell lysis buffer (20% SDS, 50% N,N-dimethylformamide, pH 3.7). After 12 h of incubation, absorbance was measured at 570 nm using a microplate reader.

With a plate confluence of 80%, cells were treated for 20 h with different doses of peptides. After this procedure, cells were washed with PBS and lysed for 45 min at 4 °C in lysis buffer. Lysed cells were centrifuged at 5000 rpm for 5 min and the supernatant was recovered. Protein concentration was determined with the BCA protein assay (Pierce, Rockford IL, USA) and samples (14 μg/lane) from the total protein fraction were analyzed by SDS-PAGE on 8% gels and further transferred to PVDF membranes (Millipore, Billerica, MA, USA). Membranes were blocked overnight at 4 °C in a solution containing Tris-buffered saline (TBS), 1% tween-20, and 5% non-fat milk. For protein detection, the following primary polyclonal antibodies were used: goat anti-β-adaptin (1:600), CALM (1:600), and rabbit anti-eps 15 (1:2500). In addition, the mouse monoclonal anti β-actin (1:500) was used. The blocked membranes were incubated with the primary antibodies for 1 h a 37 °C. After washing, horseradish peroxidase (HRP) conjugated secondary antibodies (1:5000) were added to the membrane and were incubated for 1h at 37 °C in blocking buffer. The secondary antibodies used were: donkey anti-goat IgG for β-adaptin and CALM, goat anti-rabbit IgG for eps 15, and goat anti-mouse for β-actin. Later, membranes were washed with TBS/1% tween and HRP activity was detected with the Immobilon Western kit (Millipore).

Under the different exposures using variable concentrations of helix-X, helix-Z, and the β-amyloid peptide (Aβ_25–35), EOC cells were visualized and images were taken with an Olympus IX71 microscope (100 and 400×).

Peptide samples incubated under the different conditions tested were processed using a negative staining technique and visualized using TEM. Samples (10 μL) stained with uranyl acetate solution (2% w/v) were placed on carbon-coated copper grids (400 mesh); and after incubation for 5 min at 25 °C they were dried for 10 min. TEM images were acquired using a JEM-1200EX11 JEOL microscope (70 kV).