Zinc release of Zn₇-metallothionein-3 induces fibrillar type amyloid-β aggregates

Abstract

The reactive oxygen species H₂O₂ promotes the Zn₇-metallothionein-3 induced Aβ₄₀ aggregation of fibrillar type structures via slow cysteine oxidation and Zn²⁺ release, whereas amorphous aggregates are formed by addition of Zn²⁺ to Aβ₄₀.

---

In the central nervous system, specifically the brain, relatively high concentrations of zinc are found with regional selectivity. While approximately 80% of zinc in the brain is protein-bound zinc, the remainder is concentrated in presynaptic vesicles of zinc-enriched glutamergic neurons, containing millimolar Zn²⁺ concentrations. This vesicular zinc is released into the synaptic cleft upon neuronal activity, where it can through an interaction with many different postsynaptic targets modulate their activity.¹

In Alzheimer's disease (AD) altered Zn²⁺ homeostasis is well documented. This observation along with the presence of Zn²⁺ in senile plaques, where its concentration can be as high as ∼1 mM, indicated that this metal could play an important role in this neurodegenerative disorder.^2–4 Aggregation of the peptide amyloid-β (Aβ) has been proposed to be a key event in AD. Aggregates are believed to promote a neuronal dysfunction, and later on dementia via the production of reactive oxygen species (ROS).⁵ Studies on AD animal models have also shown that genetic ablation of synaptic Zn²⁺ greatly reduces the amount of amyloid plaques.⁶In vitro studies showed that Zn²⁺ ions bind stoichiometrically to Aβ with an apparent dissociation constant of 1–10 μM at pH 7.4. (for review see ref. 7). Zn²⁺ can induce Aβ aggregation within time scale of milliseconds⁸ and can promote synaptic targeting of oligomeric Aβ.⁴ Generally, Zn²⁺ induced Aβ aggregates are rather amorphous with a lower content of fibrils and are less reactive with the amyloid marker thioflavin T (ThT) than aggregates of Aβ alone (reviewed in ref. 7 and 9). Although Zn²⁺ induces Aβ aggregation and destabilizes oligomers, it is not clear if Zn²⁺ has a protective or an exacerbating effect on the neurotoxicity of Aβ.^9,10

In the brain, a molecule involved in zinc metabolism and linked to AD is a small metalloprotein metallothionein-3 (Zn₇MT-3).^11–13 Zn₇MT-3 is mainly expressed in hippocampal glutamergic neurons that release Zn²⁺ from synaptic vesicles.¹⁴ Evidence has been provided that MT-3 occurs intra- and extracellularly.¹⁵ Structural studies revealed that the protein can bind 7 Zn²⁺ ions via its 20 cysteines, forming two metal-thiolate clusters a Zn₃–Cys₉ cluster in the N-terminal β-domain and a Zn₄–Cys₁₁ cluster in the C-terminal α-domain.^13,16

Oxidative stress is implicated in the pathogenesis and/or progression of neurodegenerative diseases. In general, ZnMTs, including the mammalian Zn₇MT-3, Zn₇MT-2 and Zn₇MT-1 isoforms, through quenching of reactive oxygen and nitrogen species (ROS and RNS) protect cells from oxidative stress. This reaction leads to cysteine oxidation/modification and Zn²⁺ release (for recent review see ref. 12). However, structural differences among Zn₇MT-3 and Zn₇MT-1/-2 are responsible for their differing reactivity and biological activity.¹⁷ Zn₇MT-3, but not Zn₇MT-1 and Zn₇MT-2 protects cultured neurons from the toxicity of Aβ.¹⁸ This protective effect has been accounted for by the metal swap between Zn₇MT-3 and Cu²⁺–Aβ, leading to the suppression of ROS production caused by the redox cycling of Cu²⁺–Aβ.¹⁹ These literature reports strongly indicate that ZnMT-3 and Aβ are co-localized near the synapses of zinc-enriched glutamergic neurons and are linked to AD and oxidative stress.

Here, we investigated if Zn²⁺ transfer from Zn₇MT-3 to Aβ₄₀ is possible in the absence and presence of H₂O₂. The term transfer was used in the sense to describe the Zn²⁺ movement from Zn₇MT-3 to Aβ₄₀ without an interaction between the two peptides. We found that Zn₇MT-3 donates Zn²⁺ to Aβ₄₀ not only in the presence of H₂O₂, but also in its absence. Moreover, we could show that Zn²⁺ transfer from Zn₇MT-3 modulates Aβ₄₀ aggregation differently than the addition of free Zn²⁺.

First, we checked by absorption spectroscopy whether Zn²⁺ may be released from Zn₇MT-3 upon addition of H₂O₂. We have followed the CysS–Zn²⁺ charge-transfer band at 230 nm over time as a function of different H₂O₂ concentrations. The observation that the charge-transfer band decreased with time in a concentration dependent manner indicated that Zn²⁺ is slowly released from the Zn²⁺-thiolate clusters in Zn₇MT-3 (Fig. 1A).

Fig. 1 (A) Time dependence of the Zn₇MT-3 absorption at 230 nm as a function of different H₂O₂ concentrations. The Cys–Zn²⁺ charge-transfer band at 230 nm was followed over 3 h. Conditions: Zn₇MT-3 20 μM, Hepes 20 mM, NaCl 20 mM, at pH 7.4. H₂O₂ concentrations used 0 μM (▽), 50 μM (▲), 100 μM (○), and 200 μM (■). (B) H₂O₂ induced Zn²⁺ release from Zn₇MT-3 monitored by complexing dye zincon. The Zn²⁺ release from Zn₇MT-3 as a function of different concentrations of H₂O₂ after a 1 day incubation. Conditions: 20 μM Zn₇MT-3, H₂O₂: 0, 15, 45 or 150 μM, Hepes 20 mM, NaCl 20 mM, pH 7.4. Prior to measurement the sample was 20 times diluted with the same buffer containing 20 μM zincon. The presented Zn²⁺ release relates to the original Zn₇MT-3 concentration of 20 μM; Inset: Time dependence Zn²⁺ release from 10 μM Zn₇MT-3 in the absence of H₂O₂ monitored trough the absorption of 20 μM zincon at 620 nm. (C) Mass spectra of apoMT-3 after incubation with different H₂O₂ concentrations. Conditions: 40 μM Zn₇MT-3 in 20 mM Hepes, 20 mM NaCl, at pH 7,4 was incubated for 1 day with 0 (black), 40 (dotted), and 200 μM (gray)) H₂O₂. Prior to analyses the samples were diluted with two volumes of H₂O–acetonitrile (50% v/v) containing 0.1% formic acid. The measurements were performed at pH 3, which induced the release of Zn²⁺. Only the 5 times charged peaks are shown. (D) Size exclusion chromatography (SEC) of Zn₇MT-3 after reaction with H₂O₂. The Zn₇MT-3 sample was incubated for 1 day with 0 (thick dotted gray), 20 (thick solid gray), 100 (thin dotted black), or 200 μM (thick solid black) H₂O₂. A freshly prepared solution of Zn₇MT-3 without H₂O₂ was used as a control (thin solid). Conditions: 20 μM Zn₇MT-3, 20 mM Hepes, 20 mM NaCl, at pH 7.4.

To obtain a more direct measure of Zn²⁺ release from Zn₇MT-3 in the presence of H₂O₂, the protein sample was incubated with increasing concentrations of H₂O₂ and the Zn²⁺ release monitored by the complexing dye zincon. Fig. 1B shows the H₂O₂ concentration dependent Zn²⁺ release from 20 μM Zn₇MT-3, obtained by following the formation of Zn²⁺–zincon complex at 620 nm over 1 day. We also observed that a small amount of Zn²⁺ (about 0.1 mole equivalent) was apparently released even in the absence of H₂O₂. To confirm this result the kinetics of Zn²⁺ release was measured directly using a higher Zn₇MT-3 concentration (10 μM) (Fig. 1B, inset). Indeed, the absorption band of the Zn²⁺–zincon complex at 620 nm increased linearly with time. A similar Zn²⁺ release from Zn₇MT-3 in the presence of increasing zincon concentrations has also been reported by Chen et al.¹⁷

To examine whether the Zn²⁺ release from Zn₇MT-3 by H₂O₂ is due to a cysteine oxidation and not its removal by zincon, a mass spectrometry (ESI-MS) was applied. We analyzed the apoMT-3 form after the Zn²⁺ depletion of the holoprotein at acidic pH. Although no deconvoluted peaks could be obtained, the results suggest that cysteine oxidation occurred by adding H₂O₂ (Fig. 1C). The apoMT-3 form has a theoretical mass of 6927.1 Da. In the absence of H₂O₂, an average mass of 6928.0 Da was obtained. However, upon addition of H₂O₂ a shift to smaller average masses was seen, in agreement with the formation of disulfide bonds (loss of two protons leads to a decrease of mass by 2 Da). In addition, a broadening of the mass peak was also observed indicating a mass heterogeneity due to the formation of a varying number of disulfide bonds in different protein molecules. Consequently, the Zn₇MT-3 structure is susceptible to oxidation by H₂O₂. Taken together, the results of zincon, ESI-MS, and the CysS–Zn²⁺ charge-transfer band absorption measurements all indicate that H₂O₂, through cysteine oxidation to disulfides, releases Zn²⁺ from Zn₇MT-3 in a concentration dependent manner.

Next we used size-exclusion chromatography (SEC) to examine whether the disulfide bridges formed upon oxidation of Zn₇MT-3 by H₂O₂ were intra- or intermolecular. In general, while in the case of intermolecular disulfide bridges the formation of dimers or higher polymers of Zn₇MT-3 should be observed, in the case of intramolecular disulfide bridges a decrease of the hydrodynamic radius of ZnMT-3, due to a more compact disulfide bridged structure would be expected.²⁰Fig. 1D shows the results of the SEC of Zn₇MT-3 after a 1 day incubation with different concentrations of H₂O₂. The freshly prepared Zn₇MT-3 eluted at about a 10.5 ml volume, a typical value for monomeric metal loaded Zn₇MT-3 with an apparent molecular mass of about 22 kDa due to its ellipsoid shape.²¹ A small amount of ZnMT-3 dimers, eluting with a volume of 9.7 ml, was also observed in the control experiments as reported.²⁰ Incubation of Zn₇MT-3 with increasing H₂O₂ concentrations resulted in two effects. First, the monomeric peak shifted to a higher elution volume. The latter effect reflects a decrease in the apparent molecular mass of the protein consistent with the increase in structure compactness brought about by a disulfide formation and Zn²⁺ release.²⁰ Second, an increase in size of the dimeric peak, in line with the formation of some intermolecular disulfide bridges, was also observed. However, since the monomeric peak even at 200 μM H₂O₂ was still much higher, this indicates that the disulfide bridges are predominantly formed intramolecularly. This conclusion is in agreement with the ESI-MS data (see above), where molecular masses consistent with predominately intramolecular disulfide bonds were obtained. Note that the apparent decrease in peak intensity seen in the elution profile of Zn₇MT-3 at a higher H₂O₂ concentration does not reflect a protein loss, but is due mainly to the decrease in the molar extinction coefficient of the CysS–Zn²⁺ charge-transfer band upon protein oxidation (see above).

The above results indicate that cysteine oxidation in Zn₇MT-3 by H₂O₂ results in a release of Zn²⁺ from the protein. Next we investigated whether the released Zn²⁺ can stimulate aggregation of the Aβ₄₀ peptide. The aggregation process was followed by turbidimetry, ThT assay, and transmission electron microscopy. The turbidimetry is a simple, but crude assay to follow the formation of aggregates. The turbidity experiments were performed at 300 nm over 100 min and are summarized in Fig. 2A. Aβ₄₀ alone showed a very small turbidity increase under the conditions used (Fig. 2A, thin solid line). In contrast, the Zn²⁺–Aβ₄₀ complex, formed upon addition of 1 mole equivalent of Zn²⁺ solution, exhibited a fast and dramatic increase in turbidity (Fig. 2A, thick solid line), in agreement with the aggregation accelerating effect of Zn²⁺ described in the literature.^9,22 Aβ₄₀ in the presence of Zn₇MT-3 and without H₂O₂ showed a small but significant turbidity increase compared to that of Aβ₄₀ alone (Fig. 2A, dotted line). However, upon addition of H₂O₂ to the latter sample the turbidity increase was much higher (Fig. 2A, dashed line), but did not reach the intensity and rapidity of aggregation of the Zn²⁺–Aβ₄₀ complex (Fig. 2A, thick solid line). The latter result indicates that the Zn²⁺ release from Zn₇MT-3 with H₂O₂ results in an intermediate turbidity between Aβ₄₀ alone and Zn²⁺–Aβ₄₀. Furthermore, in accordance with the Zn²⁺ release from Zn₇MT-3 in the absence of H₂O₂ (see above), the aggregation promoting effect has also been seen.

Fig. 2 (Upper panel) Turbidity assay: Turbidity of Aβ₄₀ was measured at 300 nm over 100 min (thin solid line), Zn-Aβ (thick solid), Aβ₄₀ and Zn₇MT-3 (dotted grey), and Aβ₄₀, Zn₇MT-3 and H₂O₂ (dashed black) Conditions: Aβ 30 μM, Zn₇MT-3 20 μM, 100 μM H₂O₂, Hepes 20 mM, NaCl 20 mM, pH 7.4. (Lower panel) Thioflavin T fluorescence: Emission spectra of ThT (10 μM), excitation at 435 nm. ThT was added after sample incubation for 1 day to Aβ₄₀ (gray dashed), Aβ and 1 equiv Zn²⁺ (gray solid), Aβ₄₀ and Zn₇MT-3 (black thick solid), Aβ₄₀, Zn₇MT-3 and H₂O₂ (black dotted), ThT alone (thin solid); Conditions: 30 μM Aβ₄₀, 20 μM Zn₇MT-3, 100 μM H₂O₂, 20 mM Hepes, 20 mM NaCl, at pH 7.4.

The measurement of turbidity only provides information about the global aggregation state. In contrast, the fluorescence dye ThT is relatively specific for amyloid structures such as that of the fibrils of Aβ. The corresponding ThT fluorescence spectra were measured after 1 day of sample incubation (Fig. 2B). In this case, Aβ₄₀ alone (Fig. 2B, gray dashed line) showed ThT fluorescence that increased upon binding of 1 mole equivalent of Zn²⁺ solution (Fig. 2B, gray solid line) in line with the Zn²⁺ amyloid promoting effect. Interestingly, although Zn²⁺ release from Zn₇MT-3 with H₂O₂ resulted in an intermediate turbidity between Aβ₄₀ alone and Zn²⁺–Aβ₄₀ (see above), the corresponding ThT fluorescence (Fig. 2B, black dotted line) revealed a value closely similar to that of the Zn²⁺–Aβ₄₀ complex prepared upon addition of Zn²⁺ solution, suggesting that the amounts of fibrilar structure are similar.¶ Moreover, since a considerable ThT fluorescence of aggregated Aβ₄₀ was also seen in the presence of Zn₇MT-3 without H₂O₂ (Fig. 2B, black solid line), this indicates that released Zn²⁺ under these conditions is also able to promote significantly the formation of Aβ₄₀ fibrils (see below).

Transmission electron microscopy was used to investigate the Aβ₄₀ aggregates formed under the different conditions (Fig. 3). The estimated amount of aggregates follows the order Zn²⁺–Aβ₄₀ > Aβ₄₀ plus Zn₇MT-3 plus H₂O₂ > Aβ₄₀ plus Zn₇MT-3 > Aβ₄₀ alone. These results parallel the turbidity measurements as expected, since turbidity measures the Aβ₄₀ aggregation. However, the content of fibrillar structures reflects more the ThT fluorescence. Thus, the solution of Aβ₄₀ and Zn₇MT-3 in the presence and absence of H₂O₂ showed the highest content of fibrillar structures, explaining why the ThT fluorescence was as high as for the Zn²⁺–Aβ₄₀ complex, although the latter form is more aggregated. This further suggests that while in the former case Aβ₄₀ aggregates slowly form substantial amounts of fibrillar structures, in the latter case fast Zn²⁺ binding to Aβ₄₀ results in more amorphous type aggregates.

Fig. 3 Transmission electron microscopy (negatively stained) of Aβ₄₀, Zn²⁺–Aβ₄₀, Aβ₄₀ plus Zn₇MT-3, and Aβ₄₀, Zn₇MT-3 plus H₂O₂ after 1 day incubation. Concentrations: 30 μM Aβ₄₀, 30 μM ZnSO₄, 20 μM Zn₇MT-3, and 100 μM H₂O₂.

In summary, the results presented above suggest that cysteine oxidation in Zn₇MT-3 in the presence and absence of H₂O₂ leads to Zn²⁺ release, which in turn can induce aggregation of Aβ₄₀. Moreover, the slow release of Zn²⁺ promotes the formation of amyloid-type fibrils in contrast to amorphous Aβ₄₀ aggregates formed by the addition of Zn²⁺. As a consequence it is the rate of cysteine oxidation in MT-3 that regulates the assembly of Aβ₄₀ aggregates and their morphology. The results are in line with the observation that zinc concentrations modulate the rate of assembly and the type of morphology of Aβ₄₀ aggregates. (See e.g.ref. 8 and 23). This effect may have biological relevance since both Aβ and Zn₇MT-3 have been linked to AD, oxidative stress and are likely to be co-localized in the brain. Zn₇MT-3 might be involved in the protection of cells not only by its capacity for ROS scavenging,²⁴ but also via Zn²⁺ release and subsequent binding to Aβ. However, at present it is not clear whether Zn²⁺ binding has a beneficial or rather detrimental effect on Aβ toxicity. Thus, the biological significance of the in vitro findings remains to be established.

Acknowledgements

This work was supported by the programme blanc NT09-488591 (Neurometals) of Agence national de la recherche (ANR) (P.F.) and Swiss National Science Foundation Grant 3100A0-111884 (M.V.). We would like to thank Cathy Claperols and Vincent Colliere (LCC Toulouse) for measuring mass spectrometry and transition electron microscopy, respectively. Dr Fatima Bousejra-El Garah (LCC) is acknowledged for measuring effect of H₂O₂ on ThT fluorescence.

References

1. C. J. Frederickson, Int. Rev. Neurobiol., 1989, 31, 145–238.
2. P. A. Adlard and A. I. Bush, J. Alzheimers Dis., 2006, 10, 145–163.
3. P. Zatta, D. Drago, S. Bolognin and S. L. Sensi, Trends Pharmacol. Sci., 2009, 30, 346–355.
4. A. Deshpande, H. Kawai, R. Metherate, C. G. Glabe and J. Busciglio, J. Neurosci., 2009, 29, 4004–4015.
5. D. G. Smith, R. Cappai and K. J. Barnham, Biochim. Biophys. Acta, Biomembr., 2007, 1768, 1976–1990.
6. J.-Y. Lee, T. B. Cole, R. D. Palmiter, S. W. Suh and J.-Y. Koh, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 7705–7710.
7. P. Faller and C. Hureau, Dalton Trans., 2009, 1080–1094.
8. D. Noy, I. Solomonov, O. Sinkevich, T. Arad, K. Kjaer and I. Sagi, J. Am. Chem. Soc., 2008, 130, 1376–1383.
9. M. P. Cuajungco and K. Y. Faget, Brain Res. Rev., 2003, 41, 44–56.
10. E. Mocchegiani, C. Bertoni-Freddari, F. Marcellini and M. Malavolta, Prog. Neurobiol., 2005, 75, 367–390.
11. J. Hidalgo, M. Aschner, P. Zatta and M. Vašák, Brain Res. Bull., 2001, 55, 133–145.
12. S. G. Bell and B. L. Vallee, ChemBioChem, 2009, 10, 55–62.
13. Z. C. Ding, Q. Zheng, B. Cai, F. Y. Ni, W. H. Yu, X. C. Teng, Y. Gao, F. Liu, D. Chen, Y. Wang, H. M. Wu, H. Z. Sun, M. J. Zhang, X. S. Tan and Z. X. Huang, J. Inorg. Biochem., 2008, 102, 1965–1972.
14. B. A. Masters, C. J. Quaife, J. C. Erickson, E. J. Kelly, G. J. Froelick, B. P. Zambrowicz, R. L. Brinster and R. D. Palmiter, J. Neurosci., 1994, 14, 5844–5857.
15. M. A. Lynes, K. Zaffuto, D. W. Unfricht, G. Marusov, J. S. Samson and X. Yin, Exp. Biol. Med. (Maywood), 2006, 231, 1548–1554.
16. M. Vašák, J. Trace Elem. Med. Biol., 2005, 19, 13–17.
17. Y. Chen, Y. Irie, W. M. Keung and W. Maret, Biochemistry, 2002, 41, 8360–8367.
18. Y. Irie and W. M. Keung, Biochem. Biophys. Res. Commun., 2001, 282, 416–420.
19. G. Meloni, V. Sonois, T. Delaine, L. Guilloreau, A. Gillet, J. Teissie, P. Faller and M. Vašák, Nat. Chem. Biol., 2008, 4, 366–372.
20. G. Meloni, P. Faller and M. Vašák, J. Biol. Chem., 2007, 282, 16068–16078.
21. P. Faller, D. W. Hasler, O. Zerbe, S. Klauser, D. R. Winge and M. Vašák, Biochemistry, 1999, 38, 10158–10167.
22. A. I. Bush, W. H. Pettingell, G. Multhaup, M. Paradis, J. P. Vonsattel, J. F. Gusella, K. Beyreuther, C. L. Masters and R. E. Tanzi, Science, 1994, 265, 1464–1467.
23. M. Innocenti, E. Salvietti, M. Guidotti, A. Casini, S. Bellandi, M. L. Foresti, C. Gabbiani, A. Pozzi, P. Zatta and L. Messori, J. Alzheimers Dis., 19, 1323–1329.
24. P. J. Thornalley and M. Vašák, Biochim. Biophys. Acta, Protein Struct. Mol. Enzymol., 1985, 827, 36–44.

---

Footnotes

† Electronic supplementary information (ESI) available: Materials and Methods. See DOI: 10.1039/c0mt00027b

‡ Present address: IPBS, 205, route de Narbonne, F-31077 Toulouse.

§ Present address: Division of Chemistry and Chemical Engineering, and Howard Hughes Medical Institute, California Institute of Technology, 114-96, Pasadena, CA 91125, USA.

¶ We verified that H₂O₂ does not affect the ThT fluorescence under the conditions used here. Thus the increase of ThT fluorescence in samples containing Aβ₄₀, Zn₇MT-3 and H₂O₂ can not be ascribed to a direct interaction of H₂O₂ with ThT. This supports the view that the increase in ThT fluorescence is due to increased Zn-release from Zn₇MT-3 upon reaction with H₂O₂.

---