Hierarchical binding of copper^II to N-truncated Aβ_4–16 peptide

Abstract

N-Truncated Aβ_4–42 displays a high binding affinity with Cu^II. A mechanistic scheme of the interactions between Aβ_4–42 and Cu^II has been proposed using a fluorescence approach. The timescales of different conversion steps were determined. This kinetic mechanism indicates the potential synaptic functions of Aβ_4–42 during neurotransmission.

---

Significance to metallomics

N-Truncated Aβ_4–x is abundant in both healthy and AD brains. Its Cu(II) binding affinity is three orders of magnitude stronger than well-known Aβ_1–42 or Aβ_1–40. Using a model peptide, Aβ_4–16, we have elucidated the reaction mechanism of Cu(II) with Aβ_4–x, crucial to understand the physiological role and toxicity of Aβ peptides. The presence of two kinetic intermediates prior to the formation of the tight ATCUN complex has implications for the potential function of Aβ_4–42 as a Cu(II) transporter during neurotransmission. The methodology used in this work may also stimulate the research of Cu(II) interactions with other intrinsically disordered proteins (IDPs).

The amyloid-β (Aβ) peptides associated with Alzheimer's Disease (AD) comprise a number of species. The “canonical” Aβ_1–42 and Aβ_1–40 peptides derived directly by proteolysis of the Amyloid Precursor Protein (APP) are complemented by N- and C-truncated species, yielded by a variety of brain proteases.¹ Among them, the N-truncated Aβ_4–42 has been reported as particularly abundant in the hippocampus and cortex of sporadic AD patients, as well as in healthy controls,^2,3 even exceeding Aβ_1–42 and Aβ_1–40.^4,5 Aβ_1–x peptides can bind Cu^II using the N-terminus and H6, H13, and His14 residues.^6–8 Hence, Aβ_1–16 has been adopted as a common model peptide in metal binding studies. K_d in the range of 0.1 nM to 1 nM at pH 7.1–7.4 was determined for Aβ_1–16 and Aβ_1–40.^9–11 The adventitious binding of Cu^II ions to Aβ_1–42/40 and the concomitant generation of reactive oxygen species (ROS) via the Cu^II/Cu^I redox pair has been proposed to be the molecular basis of oxidative stress and neuronal death in AD.¹² On the other hand, Aβ_4–x peptides bind a Cu^II ion more than three orders of magnitude more strongly (K_d = 30 fM and 6.6 fM at pH 7.4 for Aβ_4–16 and Aβ_4–9, respectively), using their N-terminal ATCUN motif spanning the Phe4, Arg5 and His6 residues. These complexes are redox-inert and do not generate significant ROS. Cu^II ion transfer from Aβ_1–16 to Aβ_4–16 occurs upon adding the latter to the Cu^IIAβ_1–16 solution.¹³ This reaction is quantitative, in agreement with the affinity difference, and fast, occurring within the sample preparation time ∼s. Such a reaction suggested that Aβ_4–42 should prevail as a Cu^II binding Aβ species in the extracellular spaces of the brain. This finding gave rise to a hypothesis that Aβ_4–42 may have a physiological role as a synaptic Cu^II scavenger during neurotransmission.¹⁴ However, Cu^II release events in glutamatergic synapses may occur on a much faster, millisecond scale. Therefore, a thorough determination of association and dissociation rate constants for the participating species is necessary to help evaluate their relevance in vivo. Such data have been obtained previously for Cu^IIAβ_1−x complexes.^15–17 Here, we studied the reaction mechanism for Cu^II binding to the model peptide Aβ_4–16 and found that the reaction follows a hierarchical fashion, going through two intermediate states and then reaching the final tight complex.

First, we studied the effect of N-truncation on the Cu^II binding kinetics. 20 nM Aβ labelled by HiLyte Fluor 488 on lysine 16 (FRHDSGYEVHHQK-HiLyte 488) was reacted with 400 nM Cu^II under various HEPES concentrations in order to obtain the HEPES-independent Cu^II binding rate constant (k_on). The results are shown in Fig. 1a. The intercept of the fitted curve (Fig. 1b) was used to determine k_on, which is 2.0(1) × 10⁸ M⁻¹ s⁻¹, 2.5 times slower than the value for Aβ_1–16.¹⁷

Fig. 1 Kinetics of Cu^II binding to Aβ_4–16. (a) Representative raw traces of Aβ (20 nM) with Cu^II (400 nM) under various concentrations of HEPES. The experiments were performed in 50 mM HEPES and 100 mM NaCl buffer solution at 298 K (pH 7.5). (b) HEPES dependence of k_on. The HEPES independent k_on value is 2.0(1) × 10⁸ M⁻¹ s⁻¹. (c) Kinetics of dissociation of Cu^IIAβ_4–16 assisted by varying concentrations of EDTA, monitored by UV-vis absorbance. The experiments were performed for 1 mM Cu^IIAβ_4–16 at 298 K (pH 7.5). (d) Empirical fit to derive the EDTA-independent k_off.

k _off was determined for the reaction of a Cu^II complex of unlabelled Aβ_4–16 with an excess of EDTA. The estimated value is ∼5 × 10⁻⁵ s⁻¹, which divided by k_on proposed here gives K_d ∼ 250 fM. EDTA is a stronger Cu^II chelator than Aβ_4–16, with a log β of 18.7, which can be recalculated into a conditional constant ^CK of 16.0 at pH 7.5.¹⁸ This value is sufficiently higher than that of Cu^IIAβ_4–16, 13.53, to assure full Cu^II transfer, as demonstrated in Fig. 1c. The reaction was carried out for a range of EDTA/peptide ratios between 2 and 120. Pseudo-1st order kinetics for the Cu^II transfer reaction was observed for all experiments. The non-linear response of k_off to EDTA required the EDTA-independent k_off value to be determined by the extrapolation of the empirical exponential fit to these data, as shown in Fig. 1d.

To gain a glimpse of a possible reaction mechanism of Cu^II binding to N-truncated Aβ_4–16, we performed binding experiments at a 1 : 1 mixing ratio of Aβ to Cu^II with increasing concentration. In such experiments, the effect of the second Cu^II binding can be ignored, as the relevant log K is as low as 6.7.¹³ The raw traces are shown in Fig. 2a. We noticed that the reaction process is becoming concentration independent after ∼2 s (results from the fit are summarized in Table S1, ESI†). Thus we infer the existence of an intramolecular process following the initial Cu^II binding.

Fig. 2 (a) Raw traces of Cu^II binding measurements with a series of 1 : 1 concentration ratio HiLyte 488 labelled Aβ and Cu^II showing the evidence of intermediate species formation. The experiment was performed in 50 mM HEPES and 100 mM NaCl buffer solution at 298 K (pH 7.5). (b) Kinetics of HiLyte 488 labelled Aβ/Cu^II interactions measured by double mixing stopped flow. Raw traces showing the change in amplitude as the delay time between mixing of an equal concentration of Cu^II and Aβ (2 μM) and addition of EDTA is increased from 50 ms to 1 min. The experiment was performed in 50 mM HEPES and 100 mM NaCl buffer solution at 298 K (pH 7.5).

Next, a double mixing stopped flow technique was employed to further explore the potential intermediate complexes formed after the initial Cu^II binding. This technique was successfully applied to probe the interconversion between component I and component II Cu^II coordination species of Aβ_1–16 and Aβ_1–40.¹⁷ 2 μM Aβ_4–16 and 2 μM Cu^II were mixed in a delay loop and after various delay times the reaction was “frozen” by adding an excess of EDTA (Fig. 2b). Taking advantage of the disparities in reactivity of different Cu^IIAβ_4–16 species with EDTA, the time evolution of the population of individual species could be resolved and analyzed, enabling us to depict details of the binding process.

As shown in Fig. 2b, the amplitude of fluorescence recovery strongly depends on the delay time, indicating that a much more inert (less reactive towards EDTA) complex (“dark” complex) formed after around 2 s. We equate this end complex, (Aβ–Cu)_D, to the very stable ATCUN-type Cu^IIAβ_4–16 complex reported previously.¹³ Furthermore, because the reaction rate is concentration independent after 2 s as mentioned above, we propose that a peptide conformational rearrangement process leading to this final complex must occur at around 2 s.

In order to describe the whole process of Cu^II binding of N-truncated Aβ_4–16, we hypothesized a reaction scheme as shown in Fig. 3a. The individual amplitudes of the two phases in Fig. 2b were determined by a global fit, which were further fitted by the scheme with KinTek software to validate it (Fig. 3b). The amplitudes indicate the amounts of two intermediates, Species I and Species II, at different reaction process stages, and could be fitted well by the predicted mechanism, with fitted rate constants listed in Table 1. A corresponding free energy landscape illustration of Cu^II binding with Aβ_4–16 is shown in Fig. 3c.

Fig. 3 (a) Reaction mechanism of Cu^II binding to Aβ_4–16 and formation of the high affinity Cu^IIAβ_4–16 complex, (Aβ–Cu)_D (Cu^II binding site shown above). (b) Fitting of the individual amplitudes at different Cu^II binding process stages of the two phases by the predicted reaction mechanism. (c) Proposed free energy landscape of Cu^II binding to Aβ_4–16. “Dark” complex refers to the very stable ATCUN-type Cu^IIAβ_4–16 complex.

Table 1 Rate constants corresponding to the mechanism scheme shown in Fig. 3a

	k +1	k −1	k +2
k value/s^−1	4.10(1)	10.34(2)	3.31(4)

---

Finally, the activation energy of the (Aβ–Cu)_D complex was determined to be 64(3) kJ mol⁻¹ (Fig. 4) by performing a series of double mixing experiments at different temperatures (raw data shown in Fig. S1, ESI†).

Fig. 4 Arrhenius plot for the switching rate constant k₊₂. The switching activation energy determined is 64(3) kJ mol⁻¹.

The chemical properties of ATCUN Cu^II complexes of Aβ_4–x peptides, such as high thermodynamic stability, absence of ROS production due to their resistance to oxidation and reduction, reluctance of copper to transfer to metallothionein-3 (MT3) and easy sequestering of Cu^II from Aβ_1–x, gave rise to a concept that Aβ_4–x peptides (full-length Aβ_4–42 and its C-truncated analogs) may serve as guardians of synaptic function, by sequestering excess Cu^II ions released during neurotransmission in glutamatergic pathways.^14,19 The key unsolved issue is how these exchange-inert complexes relay copper back to neurons to maintain the proper copper cycling. Furthermore, Cu^II-free Aβ_4–42 can be neurotoxic by forming oligomeric species.²⁰ Detailed knowledge on mechanisms of Cu^II association with and dissociation from Aβ_4–x peptides, represented here by Aβ_4–16, is thus crucial to understand the physiology and toxicity of these Aβ peptides.

The discovery of long-lived kinetic intermediates in the formation of the ATCUN complex of Aβ_4–16 is a game changer in the above considerations. The lifetimes of Species I and Species II complexes are comparable to the intervals between pulses of neurotransmitter release in glutamatergic neuronal pathways.²¹ Therefore, these complexes may well contribute to the biological activity of Aβ_4–42, and of putative short peptidic fragments generated by neprilysin cleavage, such as Aβ_4–9.^22,23 There is only one way in which four nitrogen ligands of the ATCUN motif can be arranged around the Cu^II ion, and so it is reasonable to assume that the intermediate species contain the coordinatively unsaturated Cu^II. Such species have been implicated in the reverse reaction of Cu^II dissociative transfer from Cu^IIAβ_4–16 to MT3, to explain the catalytic effect of glutamate,²⁴ but it has not been observed directly. The Species I and in particular the longer-lived Species II complex may be the actual species able to move copper around during neurotransmission. The fact that the Cu^IIAβ_1−x complex, although so much weaker, was formed 2.5 times faster, prompts further research into possible synaptic roles of Cu^II interactions with various Aβ species.

Furthermore, the observed hierarchical binding of Cu^II to Aβ_4–16 resembles the kinetics of the binding of many intrinsically disordered proteins (IDPs).²⁵ The methodology used in this study may be applicable to the fundamental understanding of the emerging “coupled binding and folding” paradigm.²⁶

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the Leverhulme Trust grant RPG-2015-345 to LY and the Biotechnology and Biosciences Research Council (UK) grant BB/R022429/1 to LY, and the National Science Centre in Poland: PRELUDIUM Grant No. 2016/21/N/NZ1/02785 and ETIUDA Grant No. 2018/28/T/NZ1/00452, to ES, and OPUS Grant No. 2018/29/B/ST4/01634 to WB. The equipment used was sponsored in part by the Centre for Preclinical Research and Technology (CePT) under award number POIG.02.02.00-14-024/08-00, a project co-sponsored by European Regional Development Fund and Innovative Economy, the National Cohesion Strategy of Poland.

Notes and references

1. D. J. Selkoe, Physiol. Rev., 2001, 81, 741–766.
2. C. L. Masters, G. Simms, N. A. Weinman, G. Multhaup, B. L. McDonald and K. Beyreuther, Proc. Natl. Acad. Sci. U. S. A., 1985, 82, 4245–4249.
3. E. Portelius, N. Bogdanovic, M. K. Gustavsson, I. Volkmann, G. Brinkmalm, H. Zetterberg, B. Winblad and K. Blennow, Acta Neuropathol., 2010, 120, 185–193.
4. G. Antonios, N. Saiepour, Y. Bouter, B. C. Richard, A. Paetau, A. Verkkoniemi-Ahola, L. Lannfelt, M. Ingelsson, G. G. Kovacs, T. Pillot, O. Wirths and T. A. Bayer, Acta Neuropathol. Commun., 2013, 1, 56.
5. T. A. Bayer and O. Wirths, Acta Neuropathol., 2014, 127, 787–801.
6. P. Dorlet, S. Gambarelli, P. Faller and C. Hureau, Angew. Chem., 2009, 121, 9437–9440 ( Angew. Chem., Int. Ed. , 2009 , 48 , 9273–9276 ).
7. B. Alies, H. Eury, C. Bijani, L. Rechignat, P. Faller and C. Hureau, Inorg. Chem., 2011, 50, 11192–11201.
8. E. Atrián-Blasco, P. Gonzalez, A. Santoro, B. Alies, P. Faller and C. Hureau, Coord. Chem. Rev., 2018, 375, 38–55.
9. B. Alies, E. Renaglia, M. Rózga, W. Bal, P. Faller and C. Hureau, Anal. Chem., 2013, 85, 1501–1508.
10. T. R. Young, A. Kirchner, A. G. Wedd and Z. Xiao, Metallomics, 2014, 6, 505–517.
11. A. Conte-Daban, V. Borghesani, S. Sayen, E. Guillon, Y. Journaux, G. Gontard, L. Lisnard and C. Hureau, Anal. Chem., 2017, 89, 2155–2162.
12. C. Cheignon, M. Jones, E. Atrián-Blasco, I. Kieffer, P. Faller, F. Collin and C. Hureau, Chem. Sci., 2017, 8, 5107–5118.
13. M. Mital, N. E. Wezynfeld, T. Frączyk, M. Z. Wiloch, U. E. Wawrzyniak, A. Bonna, C. Tumpach, K. J. Barnham, C. L. Haigh, W. Bal and S. C. Drew, Angew. Chem., 2015, 127, 10606–10610 ( Angew. Chem., Int. Ed. , 2015 , 54 , 10460–10464 ).
14. E. Stefaniak and W. Bal, Inorg. Chem., 2019, 58, 13561–13577.
15. P. Girvan, T. Miyake, X. Teng, T. Branch and L. Ying, ChemBioChem, 2016, 17, 1732–1737.
16. T. Branch, M. Barahona, C. A. Dodson and L. Ying, ACS Chem. Neurosci., 2017, 8, 1970–1979.
17. T. Branch, P. Girvan, M. Barahona and L. Ying, Angew. Chem., 2015, 127, 1243–1246 ( Angew. Chem., Int. Ed. , 2015 , 54 , 1227–1230 ).
18. J. Felcman and J. J. da Silva, Talanta, 1983, 30, 565–570.
19. N. E. Wezynfeld, E. Stefaniak, K. Stachucy, A. Drozd, D. Płonka, S. C. Drew, A. Krężel and W. Bal, Angew. Chem., 2016, 128, 8375–8378 ( Angew. Chem., Int. Ed. , 2016 , 55 , 8235–8238 ).
20. J. Dunys, A. Valverde and F. Checler, J. Biol. Chem., 2018, 293, 15419–15428.
21. W. Goch and W. Bal, PLoS One, 2017, 12, e0170749.
22. M. Mital, W. Bal, T. Frączyk and S. C. Drew, Inorg. Chem., 2018, 57, 6193–6197.
23. K. Bossak-Ahmad, M. Mital, D. Płonka, S. C. Drew and W. Bal, Inorg. Chem., 2019, 58, 932–943.
24. A. Santoro, N. Wezynfeld, E. Stefaniak, A. Pomorski, D. Płonka, A. Krezel, W. Bal and P. Faller, Chem. Commun., 2018, 54, 12634–12637.
25. K. Sugase, H. J. Dyson and P. E. Wright, Nature, 2007, 447, 1021–1025.
26. S. Gianni, J. Dogan and P. Jemth, Curr. Opin. Struct. Biol., 2016, 36, 18–24.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c9mt00299e

‡ These authors contributed equally to this work.

---