Minimal Zn²⁺ Binding Site of Amyloid-β

Abstract

Zinc-induced aggregation of amyloid-β peptide (Aβ) is a hallmark molecular feature of Alzheimer's disease. Here we provide direct thermodynamic evidence that elucidates the role of the Aβ region 6–14 as the minimal Zn²⁺ binding site wherein the ion is coordinated by His⁶, Glu¹¹, His¹³, and His¹⁴. With the help of isothermal titration calorimetry and quantum mechanics/molecular mechanics simulations, the region 11–14 was determined as the primary zinc recognition site and considered an important drug-target candidate to prevent Zn²⁺-induced aggregation of Aβ.

Alzheimer's disease, a fatal neurodegenerative disorder of the elderly, is characterized by extracellular depositions of amyloid-β peptide (Aβ) saturated with metal ions (1). It is believed that Zn²⁺ ions play a key role in pathological aggregation of Aβ and therefore affect the pathogenesis of Alzheimer's disease (2–5). The amino acid region 1–16 (Aβ(1–16)) is generally considered as the metal binding domain of Aβ (6). N-acetylated and C-amidated peptide Aβ(1–16) is soluble and stable in the presence of zinc ions under physiological conditions (7), and the three-dimensional structures of this domain in Zn²⁺-loaded and Zn²⁺-free states have been solved by NMR (8). In the Zn²⁺-Aβ(1–16) complex, zinc is tetrahedrally coordinated to His⁶, His¹³, and His¹⁴ through their N-δ1, N-ɛ2, and N-δ1 atoms, respectively, and to Glu¹¹ through its O-δ atom.

At the same time, several alternative coordination modes of zinc binding to Aβ with nonacetylated N-terminus are discussed (6,9). While there is general agreement regarding the participation of the three histidine residues, the role of Asp¹ as fourth zinc chelator instead of Glu¹¹ has been proposed (10). In addition, a pentacoordination of zinc ion by Asp¹, His⁶, Glu¹¹, His¹³, and His¹⁴ has been also suggested (11). However, in contrast to the data published by Zirah et al. (8), none of the alternative complexes has been defined structurally. Thus, the exact mode of zinc coordination by Aβ deserves further investigation.

In this work, we characterize the thermodynamics of the binding of Zn²⁺ to Aβ fragments and Aβ(1–16) mutants by isothermal titration calorimetry (ITC) in order to determine the minimal Zn²⁺-binding site of Aβ under physiological conditions. We also use the quantum mechanics/molecular mechanics (QM/MM) method to model Zn²⁺ recognition by Aβ.

Thermodynamic parameters of Zn²⁺ binding to Aβ fragments and Aβ(1–16) mutants obtained by ITC (Fig. 1 and Fig. S1 in the Supporting Material) are presented in Table 1. It can be seen that acetylation of N-terminus does not dramatically change Zn²⁺ binding to Aβ(1–16).

Figure 1.

ITC titration curves (upper panel) and binding isotherms (lower panel) for zinc interactions with Aβ(1–16), Aβ(6–14), and Aβ(6–13) (black, red, and blue curves, respectively) at 25°C in 50 mM Tris buffer, pH 7.3.

Table 1.

Thermodynamic parameters of Zn²⁺ ions binding to Aβ fragments and Aβ(1–16) mutants obtained by ITC at 25°C in 50 mM Tris buffer, pH 7.3

Peptide∗	N†	Ka§	ΔH†	TΔS
		(M−1 × 104)	(kcal M−1)	(kcal M−1)
Aβ(1–16)	1.1	1.80	−4.0	1.8
NH2-Aβ(1–16)	1.1	3.90	−5.4	0.9
Aβ(6–14)	1.0	8.60	−9.2	−2.5
Aβ(6–13)	0.6	0.34¶	−8.1	−3.3
Aβ(7–14)	0.9	0.17	−7.9	−3.5
Aβ(11–14)	0.5	0.27¶	−2.6	2.1
Aβ(1–5)	No binding
Aβ(1–16) H6R	0.8	0.25	−6.4	−1.8
Aβ(1–16) E11A	1.0	0.54	−6.9	−1.8

All experiments were carried out on an iTC₂₀₀ instrument (MicroCal, Northampton, MA).

^∗All peptides were purchased from Biopeptide (San Diego, CA). The N- and C-termini of each peptide were protected with acetyl and amide, respectively; peptide NH₂-Aβ(1–16) was protected at the C-terminus with amide.

^†Standard deviation did not exceed ±10%.

^§Standard deviation did not exceed ±20%.

^¶Due to peptide dimerization, K_a dimension is equal to M⁻².

The truncation of five amino acids from the N-terminal part and two from the C-terminus of Aβ(1–16) leads to the fivefold increase in its affinity to Zn²⁺ (Fig. 1, Table 1). This could be explained by the formation of a hydrogen bond between His⁶ and either Asp¹ or Glu³ in Aβ(1–16) in the absence of zinc (Fig. S2). At the same time, titration of the Aβ(1–5) fragment with Zn²⁺ ions did not produce any heat. It is important to note that angiotensin-converting enzyme proteolytically cleaves this fragment from Aβ (12). Thus, angiotensin-converting enzyme can increase zinc affinity to Aβ and act as a modulator of zinc binding. Visible aggregation of Aβ(6–14) peptide was observed in the calorimetrical cell upon interaction with zinc. However, this aggregation does not affect stoichiometry of Zn²⁺ binding (Table 1), suggesting that formation of oligomers occurs not via Zn²⁺ ions, but through interpeptide interactions of Zn²⁺-loaded Aβ(6–14) in the same manner as for Aβ which first binds to zinc and then is conformationally driven to aggregates (13). Further removing of one of the terminal histidines from Aβ(6–14) results in the remarkable drop in association constant K_a from 8.6 × 10⁴ M⁻¹ to 1.7 × 10³ M⁻¹ for peptide Aβ(7–14) and to 3.4 × 10³ M⁻¹ for peptide Aβ(6–13).

These data indicate that all Zn²⁺-chelating amino acids are located within the Aβ(6–14) peptide and that the Aβ (1–5) segment of Aβ is not necessary for Zn²⁺ binding. Amino acids located between Asp⁷ and Tyr¹⁰ of Aβ do not participate in Zn²⁺ binding, because their removal (Aβ(7–14) → Aβ(11–14)) only slightly changes the affinity of the peptide to zinc. Substitution of Glu¹¹ to Ala as well as His⁶ to Arg in Aβ(1–16) causes strong decrease of Zn²⁺ affinity (Table 1). These data clearly demonstrate that His⁶, Glu¹¹, and His¹⁴ are necessary for Zn²⁺ coordination, which is in good agreement with NMR structural data (8).

Zinc binding to all studied Aβ fragments and Aβ(1–16) mutants was enthalpy-driven (Table 1). Contrary to the Aβ(1–16) domain, Zn²⁺ binding to Aβ(6–14), Aβ(7–14), and Aβ(6–13) peptides was entropically unfavorable. This is explained by burying hydrophobic surfaces upon Zn²⁺ binding to Aβ(1–16), which is in line with the structural data (8). Stoichiometry of Zn²⁺ binding to Aβ(6–13) and Aβ(11–14) fragments is close to 0.5 (Table 1), indicating dimer formation via Zn²⁺ ion which is in a good agreement with a recent simulation study (14). High positive entropy of Zn²⁺ binding to Aβ(11–14) dimer is explained by formation of hydrophobic contacts between the two subunits.

This is also confirmed by an NMR study of the Zn²⁺-Aβ(11–14) dimer. ¹H NMR spectra (Fig. S3) demonstrate that resonance peaks of both valine methyl groups are strongly upshifted upon Zn²⁺ binding to Aβ(11–14) (from 0.83 and 0.75 to 0.45 and 0.69 ppm), due to the proximity of the zinc ion. The peak that displays the stronger shift undergoes significant broadening, also indicating that the corresponding methyl group is very close to Zn²⁺. The coordination of Zn²⁺ by two peptide molecules resulting in the formation of a symmetrical complex is in good agreement with NMR data as well. In such a complex, valine methyl groups within the immediate neighborhood of Zn²⁺ ion are oriented inward, which means that they are not exposed considerably to the solution.

Thus, we propose that the tetrapeptide Aβ(11–14) containing three Zn²⁺ chelating amino acids functions as a Zn²⁺ recognition site. In line with this hypothesis, NMR data demonstrate that this site has a definite structure in the absence of Zn²⁺ (8). We suppose that Zn²⁺ capture by the structured EVHH region is the initial step in the formation of the Zn²⁺-Aβ(1–16) complex.

The dimerization of the Aβ(11–14) peptide upon Zn²⁺ binding makes it impossible to determine the thermodynamic parameters of binding for the monomeric peptide, i.e., to reveal the role of this fragment in the mechanism of Zn²⁺ recognition by Aβ in more detail. Therefore, we applied an approach similar to that described by Furlan and La Penna (15). Theoretical calculations of the stability of a transitional EVHH complex with zinc were performed. QM/MM simulations were applied to analyze Zn²⁺ binding by Aβ(11–14), Aβ(6–14), and Aβ(1–16) peptides. Starting conformation for each peptide complexed with Zn²⁺ was derived from the Aβ(1–16) structure (PDB ID: 1ZE9).

Simulations showed that both Aβ(6–14) and Aβ(1–16) peptides where Zn²⁺ was chelated by His⁶, Glu¹¹, His¹³, and His¹⁴ kept stable tetrahedral coordination of zinc during 8 ps of the QM/MM simulation trajectory. In the case of Aβ(6–14) energy, the QM part (see the Supporting Material for details) was slightly smaller because His¹³ is better oriented in the complex (Table 2). For Aβ(11–14), fast (1-ps) formation of the tetrahedral Zn²⁺ coordination environment (Table 2) with a water molecule as the fourth chelator was observed (see Movie S1 in the Supporting Material). In the longer (8-ps) simulation, this water molecule was stable in terms of orientation and position. Summarizing our simulation results, we conclude that initially Zn²⁺ is recognized and captured by the EVHH region of Aβ and temporarily coordinated by water as the fourth chelator (Fig. 2 A). This state exists until His⁶ comes close to Zn²⁺ due to thermal fluctuations and replaces the water molecule in the zinc coordination environment, resulting in the formation of the final complex (Fig. 2 B).

Table 2.

Energy and geometry values of zinc coordination by Aβ(1–16), Aβ(6–14), and Aβ(11–14) peptides obtained by QM/MM simulations

	Aβ(1–16)	Aβ(6–14)	Aβ(11–14)
Energy of QM subsystem, kJ/mol
	−618571 ± 27	−618621 ± 64	NA
Distance, nm
Zn2+ → His6 (Nδ)	0.202 ± 0.006	0.203 ± 0.005	0.206 ± 0.010
Zn2+ → H2O (O)	NA	NA	0.206 ± 0.010
Zn2+ → Glu11 (Oδ)	0.220 ± 0.014	0.226 ± 0.015	0.200 ± 0.007
Zn2+ → His13 (Nɛ)	0.207 ± 0.006	0.207 ± 0.006	0.202 ± 0.007
Zn2+ → His14 (Nδ)	0.209 ± 0.006	0.209 ± 0.006	0.199 ± 0.006
Angle, degrees
His6-Zn2+-Glu11	96.5 ± 4.6	95.6 ± 5.2	NA
His6-Zn2+-His13	131.0 ± 3.7	134.2 ± 4.2	NA
His6-Zn2+-His14	110.8 ± 4.2	110.4 ± 4.6	NA
Glu11-Zn2+-His13	85.2 ± 4.5	83.0 ± 3.2	98.1 ± 6.1
Glu11-Zn2+-His14	96.3 ± 4.6	96.9 ± 5.2	112.6 ± 7.0
His13-Zn2+-His14	117.0 ± 5.2	114.0 ± 4.2	116.5 ± 5.7
H2O-Zn2+-Glu11	NA	NA	97.5 ± 6.9
H2O-Zn2+-His13	NA	NA	111.9 ± 8.0
H2O-Zn2+-His14	NA	NA	115.5 ± 7.9

NA: not applicable.

Figure 2.

Schema of Zn²⁺ ion recognition by the Aβ binding site 6–14 according to QM/MM calculations performed with the GROMACS/CPMD package (http://www.tougaloo.edu/research/qmmm/index.htm) based on PDB ID: 1ZE9. (A) structure of the intermediate Zn²⁺-Aβ complex with a water molecule as the fourth chelator; (B) structure of the final Zn²⁺-Aβ complex. See also Movie S2 in the Supporting Material.

In summary, we used ITC to characterize the interactions of Zn²⁺with fragments of native Aβ and Aβ(1–16) mutants. The amino acid region 6–14 of Aβ was determined as the minimal Zn²⁺-binding site wherein the ion is coordinated by His⁶, Glu¹¹, His¹³, and His¹⁴. Both ITC and QM/MM showed that three of four residues from the Aβ region 11–14 (EVHH) contribute to zinc binding, and that this tetrapeptide readily forms dimers linked through a zinc ion, similarly to a model proposed for Aβ aggregation in a recent molecular dynamics study (14).

These data allow us to consider Aβ(11–14) tetrapeptide as a primary Zn²⁺-recognition site of Aβ and an important drug target candidate to prevent Zn²⁺-induced aggregation of Aβ.