The successful immobilization of the Aβ peptide on the sensor chip surface was tested by injecting a Zn²⁺ ion solution into the flow chamber and pumping it across the Aβ-immobilized sensing surface. In our experiment, the sample loop volume is 20 μL, the flowing rate is 10 μL/min, so that the Zn²⁺ ions start to flow onto the immobilized Aβ_1–28 at about 80 s, the flowing of metal ions ends at about 200 s; the difference in the baseline SPR angles before and after the Zn²⁺ ion solution injection (termed as the SPR dip shift Δθ) is approximately 0.00246° (Figure 1a), considering the sophisticated interaction between Aβ peptide and metal ions, the detection of the SPR dip shift Δθ is selected at 550 s.

To confirm that the net change is induced by the metal ions, we conducted a control experiment in which Zn²⁺ ions flowed over an ethanolamine-blocked surface without the immobilization of Aβ peptide. Figure 1b shows that the SPR signal returned to its baseline within a short time. Additionally, the recovery of the original baseline in curve b suggests that the blocking process is efficient because it has been reported that alkanethiol SAM containing carboxylic acid can be used for SPR measurements of heavy metal ions [23]. Furthermore, the conversion of the negatively charged carboxyl groups on the MUA SAM to neutral amide groups during the peptide immobilization and blocking process could reduce the undesirable electrostatic attraction to metal ions in solution.

Figure 2 shows the sensorgrams of various Zn²⁺ concentrations flowing over the Aβ_1–28 sensor chip. In these experiments, EDTA (10 mM) was used to regenerate the surface. Therefore, a single chip can be used repeatedly for multiple samples. As shown in Figure 2, increasing Zn²⁺ concentrations results in a greater SPR dip shift. When 600 μM Zn²⁺ flowed over the Aβ_1–28-covered SPR sensor, the metal ions caused a net change of 0.0075°, which can be attributed to the binding of Zn²⁺ ions to Aβ_1–28 resulting in conformational changes of the peptide molecules.

The SPR angle shift is directly correlated with the Zn²⁺ concentration ([Zn²⁺]). In Figure 3, the calibration curve generated from the sensor response to a series of different Zn²⁺ concentrations is shown. The calibration curve contains two regions; the first is from 50 to 300 μM, and the second from 600 to 800 μM. Linear regression analysis of both regions yielded the following equations:

The slope of the linear regression in the second region is higher than that in the first region, which reveals that the SPR angle shift elicited by Zn²⁺ ions is greater with increasing Zn²⁺ ion concentration. These results suggest that the Zn²⁺-induced Aβ_1-28 conformation change is concentration dependent. This finding is in agreement with a previous report indicating that Aβ undergoes a conformational change from a random coil to a regular secondary structure in the presence of Zn²⁺ ions and forms stable 1:1 and 1:2 (peptide/zinc) complexes [8]. A possible explanation for this phenomenon is that the Aβ_1–28 binds Zn²⁺ intramolecularly at low concentrations. When the Zn²⁺ ion concentration is sufficiently high, the excess Zn²⁺ ions may begin to bind with Aβ_1–28 intermolecularly, and the Zn²⁺ ions behave like a bridge connecting two adjacent Aβ_1–28 molecules. Our finding is supported by previous studies that have shown that Zn²⁺ can coordinate to Aβ in an intra- and inter-peptide mode [24–26].

A similar trend was not found for the interactions between Cu²⁺ and Aβ_1–28. The calibration curve depicted in Figure 4 only contains one linear region ranging from 100 to 600 μM. The linear regression equation with a correlation coefficient of R² = 0.99 suggests a linear relationship between the SPR angle shift and the Cu²⁺ concentration.

Compared with Zn²⁺-Aβ complexes having stoichiometry ranging from 1:1 to 3:1 [27–29], most studies have demonstrated that the Aβ-peptide forms a 1:1 complex with Cu²⁺ [25,30,31]. NMR studies have demonstrated that Aβ_1–28 forms a 1:1 complex with the Cu²⁺ ion via histidine residues [30]. We interpret these results to different coordination modes. Unlike Zn²⁺, which can bind to Aβ in an intra- and inter-peptide coordination mode, Cu²⁺ is primarily involved in intra-peptide binding in a fairly closed structure that protects the metal from further interactions. In addition, it has been reported that under physiological conditions, the coordination of the Cu²⁺ equivalent to the Aβ peptides leads to a mononuclear complex Cu₁(Aβ)₁ and is unlikely to form a Cu₂(Aβ)₁ complex [32].

Aβ_1–40 and Aβ_1–42 are the most prevalent in vivo Aβ forms. In particular, Aβ_1–42 has a high propensity to self-assemble and deposit in senile plaques and is highly toxic to neurons [33]. While great progress has been achieved on the coordination chemistry of Zn²⁺ and Cu²⁺ with truncated Aβ_1–16 and Aβ_1–28, it is important to understand whether these peptides, which are missing a large part of the hydrophobic C-terminal residues, coordinate metal ions differently than the full-length peptides. Therefore, to better understand the mechanisms of AD, we continued to study the interactions of Zn²⁺ and Cu²⁺ with Aβ_1–42 using the FI-SPR biosensor.

Figure 5 depicts SPR sensorgrams obtained from the sensor response to various Zn²⁺ concentrations. In contrast to the calibration curve of the Zn²⁺ interactions with Aβ_1–28 that contains two regions, the calibration plot of the Zn interactions with Aβ_1–42 only contains one region; the linear regression equation is as follows:

Aβ_1–42 peptides have 14 additional residues in the hydrophobic tail compared to Aβ_1–28, which makes Aβ_1–42 more prone to aggregation than the truncated Aβ_1–28. A possible explanation for the decreased stoichiometric binding of Aβ_1–42 compared to Aβ_1–28 may be the interference of the hydrophobic tail in Aβ_1–42 with the Zn²⁺ binding site. Our data are in agreement with a previous report indicating that Zn and Cu form a monomeric complex with Aβ_1–42 [34].

Similar results were obtained for the Cu-Aβ_1–42 interaction. As shown in Figure 6, the SPR angle shift increased with an increase of Cu²⁺ bound to the Aβ_1–42-immobilized sensor chip. The calibration curve generated from the sensor exposed to different concentrations of Cu²⁺ is shown in the Figure 6 inset. A strong correlation coefficient (R² = 0.99) was obtained for the linear regression equation calculated using Cu²⁺ concentrations ranging from 50 to 400 μM.

Human amyloid-β peptide (1–28) (Aβ_1–28) and human amyloid-β peptide (1–42) (Aβ_1–42) were obtained from GL Biochem Ltd. (Shanghai, China). 1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC), N-hydroxysulfosuccinimide (NHS), 11-mercaptoundecanoic acid (MUA), ARC-grade dimethyl sulfoxide (DMSO, 99%) and ethanolamine were purchased from J & K Chemical Ltd. K₂HPO₄·3H₂O, KH₂PO₄, NaCl, ZnCl₂ and CuCl₂ were all of AR grade and purchased from Beijing Chemical Reagent Co. (Beijing, China). Water with a resistivity of 18.25 MΩ·cm⁻¹ was collected from a Millipore Simplicity 185 system. EDC (0.4 M) and NHS (0.1 M) were prepared in a pH 7.4 PBS buffer (10.0 mM K₂HPO₄·3H₂O and KH₂PO₄ prepared in 1 mM NaCl). MUA (4 mM) was prepared in pure ethanol. Stock solutions of 2.0 mM ZnCl₂ and CuCl₂ were prepared in water and then diluted to the desired concentrations.

Uniform and nonaggregated monomeric Aβ_1–28 and Aβ_1–42 peptides were prepared as previously described [35]. Briefly, 0.1 mM of the peptide was dissolved in DMSO, which disrupts the β-sheet structure and renders Aβ monomeric [36]. The freshly dissolved monomeric Aβ was diluted with PBS buffer (pH 7.4) to a concentration of 10 μM.

BK7 glass cover slides (Fisher) were cleaned with a piranha solution of 70% concentrated H₂SO₄ and 30% H₂O₂ (7:3, v/v) at 80 °C for 30 min. Upon cooling to room temperature, the glass slides were rinsed thoroughly with deionized water. After drying with N₂, each glass slide was coated with a 2 nm chromium layer and then covered with 50 nm of gold film using a sputter coater (Model 108, Kert J. Lester Inc., Clariton, PA, USA).

Aβ peptide immobilization was achieved using covalent bonding mediated by a chemical reaction. The N-terminus of the Aβ peptide was reacted with the functional group of the SAM of MUA on the surface of the Au film (as depicted in Figure 7). Briefly, the MUA gold chip was treated with a mixture of 0.4 M EDC and 0.1 M NHS (1:1) for 3 h to ensure that the carboxyl group of the SAM reacted fully with the EDC and NHS. Then, a freshly prepared 10 μM Aβ solution in PBS (pH 7.4) was reacted with the NHS-activated surface for 2 h. Finally, ethanolamine (1 M, pH 8.5) was used to block the remaining activated surface groups. The resulting film was either used immediately or stored at 4 °C for future use.

SPR measurements were conducted with a custom-built flow injection-SPR equipped with a bicell detector as previously described [22]. The SPR instrument recorded the reflected light on two photodetectors (A and B). The differential (A − B) and sum (A + B) signals were then detected by a PCI-1371 interface card (Advantech, Taiwan) controlled by the Labview program. The resonance angles from the biosensor were measured by the division of the differential and sum signals, (A − B)/(A + B). The inlet of the flow cell was connected to a six-port valve. For each measurement, the sample solution was injected into a 20-μL loop with a microsyringe (Hamilton) and subsequently delivered to the flow cell at a flow rate of 10 μL/min, using a syringe pump (KDS100, KD Scientific Inc., Holliston, MA, USA).