On the ability of CuAβ_1-x peptides to form ternary complexes: Neurotransmitter glutamate is a competitor while not a ternary partner

Highlights

• The ternary complex formation by amyloid beta (Aβ) peptides is an important issue.

• We studied interactions of Cu(II) complex of Aβ_1–16 with Glu by several techniques.

• The absence of ternary Aβ_1–16Glu complexes was demonstrated.

• Glu is able to compete for Cu(II) with Aβ_1–16 under biologically relevant conditions.

• We discuss general rules for formation of ternary complexes of Aβ_1–16 and Cu(II)

Abstract

In the light of conflicting reports on the ability of copper(II) complexes of amyloid beta (Aβ) peptides to form ternary complexes with small molecules co-present in the biological milieu, we performed a study of coordination equilibria in the system containing Cu(II) ions, the Aβ_1–16 peptide, glutamic acid and 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (2-[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfonic acid, HEPES) buffer. Using potentiometry, isothermal titration calorimetry (ITC), UV–visible spectroscopy and EPR, we concluded that glutamic acid was not able to form such a ternary complex, but can efficiently compete for the Cu(II) ion with the Aβ peptide at Glu concentrations relevant for the synaptic cleft. We also found that the literature constants for Cu(II) complexes with Glu were overestimated, but this effect was partially compensated by the formation of a ternary Cu(Glu)(HEPES) complex. Our results indicate that small molecules co-present with Cu(II) ions and Aβ peptides in the synaptic cleft are not very likely to enhance Cu(II)/Aβ interactions, but instead should be considered as a Cu(II) buffering system that may help prevent these interactions and participate in Cu(II) clearance from the synaptic cleft.

Introduction

The Cu(II) interactions with Aβ peptides, predominantly Aβ_1–42 and Aβ_1–40, have been implicated in the etiology of Alzheimer's Disease (AD) [1], [2], [3], [4]. Major lines of evidence described in the literature include association of Cu(II) ions with amyloid plaques composed of Aβ peptides isolated from post-mortem AD brains, enhancement of Aβ aggregation by Cu(II) ions, and support of Cu(II)/Cu(I) redox cycling by the Aβ peptides, yielding deleterious reactive oxygen species. Monomeric Aβ peptides bind one Cu(II) ion with a conditional stability constant at pH 7.4 (^CK_7.4) of 10¹⁰ M^− 1, and another at least 100 times weaker [5], [6]. These relatively low stability constants, determined for both the short model peptide Aβ_1–16, as well as the physiologically relevant Aβ_1–40 species, may be considered to weaken the proposal associating the Aβ neurotoxicity with Cu(II) binding, because of the presence of stronger Cu(II) ligands in the brain. Human serum albumin (HSA), which binds one Cu(II) ion with ^CK_7.4 of 10¹² M^− 1 [7], is abundant in the cerebrospinal fluid at a micromolar level [8]. Thus, a higher conditional stability constant for Cu(II) binding, compared with Aβ_1-x peptides, co-localization with these peptides, and a high abundance of HSA suggest that this protein could effectively compete for Cu(II) with Aβ_1-x peptides. Indeed, this protein was shown to withdraw Cu(II) from Aβ_1-x peptides directly and quickly, and was suggested by several research groups, including us, to be a “brain guardian” preventing Aβ-associated copper toxicity [9], [10], [11], [12]. Recently, we combined our experimental results with previous analytical reports to show that another Aβ species, namely Aβ_4-x (x = 16 is the model peptide, x = 42 is a dominant brain species) binds one Cu(II) ion with ^CK_7.4 of 3.4 × 10¹³ M^− 1, and withdraws Cu(II) from the Aβ_1-x species quantitatively within the time required for sample mixing [13]. Taking into account numerous reports that Aβ_4-x species are at least as abundant as the Aβ_1-x species in both healthy and AD brains [14], [15], [16], [17], and that the main Cu(II) complex of the Aβ_4-x peptide is redox inactive, one can wonder whether the Cu(II)-mediated Aβ toxicity may ever occur in vivo.

One way of resolving this dilemma is by considering that the aggregated Aβ_1-x peptides could bind Cu(II) more strongly than the monomeric peptide. This idea was tested experimentally by two research groups, but conflicting results were obtained (i.e. no enhancement vs. a significant and progressive enhancement) [18], [19]. However, the highest enhancement of binding reported [19] still would not overcome the affinity characteristic of the Aβ_4-x peptides [13].

The formation of ternary complexes could be another way of enhancing Cu(II) binding to Aβ_1-x peptides. The susceptibility of model peptide Aβ_1–16 to formation of ternary complexes with small molecules and common buffers was postulated in our early paper [20]. Further research did not yield evidence for such abilities of HEPES buffer, but a recent report indicated the formation of a ternary Tris complex (although the reaction was studied only at pH 4.0) [21]. In our recent paper we demonstrated that Aβ_1–16 was able to form a ternary Cu(II) complex with 2-[(dimethylamino)methyl]-8-hydroxyquinoline, an analog of the experimental drug PBT2 (5,7-dichloro-2-[(dimethylamino)methyl]quinolin-8-ol) which was designed to compete Cu(II) from the Aβ complex [22]. Consideration of the most sui candidates for a ternary ligand in vivo led us to propose glutamic acid (Glu), since AD pathology initially affects glutamatergic synapses and the concentrations of Glu in these synapses during neurotransmission may exceed 1 mM [23]. We therefore sought to identify ternary complexes in the Cu(II)/Aβ_1–16/Glu system.