Condition-Dependent Coordination and Peroxidase Activity of Hemin-Aβ Complexes

Abstract

The peroxidase activity of hemin-peptide complexes remains a potential factor in oxidative damage relevant to neurodegeneration. Here, we present the effect of temperature, ionic strength, and pH relevant to pathophysiological conditions on the dynamic equilibrium between high-spin and low-spin hemin-Aβ₄₀ constructs. This influence on peroxidase activity was also demonstrated using 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) and dopamine (DA) oxidation rate analyses with increasing ratios of Aβ₁₆ and Aβ₄₀ (up to 100 equivalents). Interaction and reactivity studies of aggregated Aβ₄₀-hemin revealed enhanced peroxidase activity versus hemin alone. Comparison of the results obtained using Aβ₁₆ and Aβ₄₀ amyloid beta peptides revealed marked differences and provide insight into the potential effects of hemin-Aβ on neurological disease progression.

1. Introduction

Central nervous system (CNS) neurodegeneration is characterized by the dysfunction and death of neurons in the brain and spinal cord. Recent efforts focused on elucidating the pathophysiological origins of CNS neurodegeneration, coupled with clinical failures on clearing amyloid beta constructs, suggests the need for an in-depth understanding of the etiology in order to design better treatment protocols [1,2,3]. Within this paradigm considerable effort remains focused on determining the roles that metal ions may play in the structural modification of protein aggregates, as well as aberrant oxidative and nitrative modification of various cellular components (e.g., DNA, cell membrane, proteins, and small molecules) [4,5,6,7].

The dysfunction of regulatory ferric heme (hemin) pathways present in neurodegenerative diseases, compounded by the interaction of hemin with amyloid beta (Aβ) and attendant peroxidase activity, has led to the proposal that hemin plays a role in disease progression [8,9,10]. Hemin has also been found to suppress Aβ-induced inflammatory activation of astrocytes and may retard other amyloid clearance pathways, further complicating etiological analyses [11,12]. A number of elegant studies by the groups of Casella [13], Atamna [14,15,16] Gao [17], Ghosh and Dey [18,19,20], Flemmig [21], Lu and Peng [22] and Imhof [23,24] have demonstrated the interaction between hemin and Aβ peptides (i.e., Aβ₁₆, Aβ₄₀, Aβ₄₂, and mutant Aβ) with an increase in peroxidase activity [9]. Previous work has also led to the suggestion that at high concentrations of Aβ and hemin or high ratios (ca. 8 equivalents) of Aβ to hemin, a cytochrome b-type species predominates and can attenuate reactivity [20]. However, yet to be catalogued fully are the precise differences in coordination and peroxidase activity observed with Aβ₄₀, a major amyloid species in diseased brains [25], at physiological temperature (37 °C), pH relevant to neuroinflammation (acidic pH), at high Aβ (i.e., 20 to 100 equivalents) to hemin ratios, and with aggregated Aβ. Here, we report a detailed condition-dependent modulation of the coordination environment and reactivity of hemin-Aβ constructs. As described below, under conditions designed to mimic those associated with neuroinflammation, such as at acidic pH (i.e., pH 6.3) [26,27] and at high ratios of Aβ-to-hemin that are potentially reflective of local concentrations in neuroinflammation or after traumatic brain injury (TBI) [21,28,29,30], the peroxidase activity of hemin-Aβ₄₀ and Aβ₁₆ complexes were significantly enhanced, as demonstrated via the oxidation of 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) to ABTS radical cation and the oxidation of dopamine to dopaminochrome.

2. Results and Discussion

The interaction between hemin (5 μM), which is present as a mixture of dimeric and monomeric species in aqueous media [31,32], and Aβ₄₀ was studied at physiologically relevant pH 7.4 in low ionic strength (5 mM) phosphate buffer at 23 °C. Upon adding up to two equivalents of Aβ₄₀, the hemin Soret band (λ_max = 390 nm) was observed to undergo a decrease in molar absorptivity along with a bathochromic shift in the maximum to λ_max = 394 nm (Figure 1).

These changes in the optical signature are ascribed to the formation of an intermediate species recognized as containing a high-spin, five coordinate 1:1 Aβ₄₀: hemin species [13]. In fact, Aβ₄₀ can coordinate both the monomeric and dimeric forms of hemin giving two species that are not spectrophotometrically distinguishable (Supplementary Materials, Scheme S1). As the concentrations of Aβ₄₀ are raised, a progressive increase in the molar absorptivity is seen in the UV-vis spectrum with an attendant bathochromic shift to λ_max = 414 nm and the appearance of peaks at λ = 532 and 560 nm also being seen. At such concentrations, the low-spin, six coordinated 2:1 Aβ₄₀: hemin species predominates [13,18]. The calculated binding constant for the 2:1 Aβ₄₀: hemin complex is log β₂ = 10.84 ± 0.02 (

Table 1. Equilibrium constants for hemin binding to Aβ₄₀.

Complex a	log K1	log K2 b	log β2
[hemin(Aβ40)2]	6.08 ± 0.02	4.77	10.84 ± 0.02
[hemin(Aβ40)2] c	5.60 ± 0.01	4.20	9.80 ± 0.01
[hemin(Aβ40)2] d	4.76 ± 0.03	4.30	9.05 ± 0.01
[hemin(Aβ40)2] e	5.30 ± 0.02	4.47	9.77 ± 0.04
[hemin(Aβ40)2] f	7.30 ± 0.02	4.02	11.32 ± 0.16
[hemin(Aβ16)2] g	4.80 ± 0.02	4.03	8.82 ± 0.02

^a Standard conditions: hemin (5 μM) in pH 7.4 phosphate buffer (5 mM) at 23 °C. ^b Obtained as the difference between log β₂ and log K₁. ^c pH 6.89. ^d pH 6.3. ^e 50 mM phosphate buffer pH 7.4. ^f 37 °C. ^g 50 mM PBS, pH 7.5, 27 °C; [13].

). By increasing the ionic strength (50 mM phosphate buffer), again at pH 7.4 and 23 °C, similar spectral features were observed (Supplementary Materials, Figure S1). Under these conditions, the calculated binding constant for the 1:1 complex was higher (log K₁ = 7.30 ± 0.02) while that of the 2:1 complex was lower, log β₂ = 9.77 ± 0.04 (Δlog β₂ = 1.07).

Neuroinflammation with attendant acidosis is an established biomarker in neurodegenerative disease [26,27]. Titrations of Aβ₄₀ at 23 °C in pH 6.89 and 6.3 phosphate buffer (5 mM) resulted in decreased binding constants of log β₂ = 9.80 ± 0.01 and 9.05 ± 0.01, respectively (Supplementary Materials, Figure S2). Notably, a significant increase in the formation of the 1:1 Aβ₄₀: hemin complex, Δlog K₁ = 1.22, and 2:1 Aβ₄₀: hemin complex, Δlog β₂ = 0.48, was seen upon increasing the temperature of the standard conditions (pH 7.4, 5 mM phosphate buffer) to 37 °C (K₁ = 7.30 ± 0.02, log β₂ = 11.32 ± 0.16; Supplementary Materials, Figure S3). In relation to other axial ligands, the log β₂ value for 50 mM phosphate buffer (pH 7.4) at ambient temperature is still larger than Aβ₁₆ (8.82 ± 0.02) [13], L-histidine (2.95) and other histidine-containing peptides (3.91 to 7.67) [33], or the tau repeat peptide I (log K₁ = 4.96 ± 0.01 [34]; log β₂ could not be determined due to small log K₂ values) studied under comparable conditions. This data is also in agreement with previously reported values for Aβ_40/42 [14,15,16,17,18,35].

The dynamic equilibrium between low-spin, six coordinate hemin(Aβ₄₀)₂ (1:2) and high-spin, five coordinate hemin-Aβ₄₀ (1:1) was also probed qualitatively by decreasing the temperature from 37 to 5 °C at 5, 10, 20, and 50 equiv. of Aβ₄₀ (Supplementary Materials, Figure S4). Analogous to the previous work on Aβ₁₆-hemin complexes at various temperatures by our group, upon decreasing the temperature, the equilibrium increasingly favours the 2:1 Aβ₄₀: hemin complex [13]. At 5 and 10 equiv., upon decreasing the temperature from 37 °C to 4 °C, reversion from the 2:1 Aβ₄₀: hemin to a mixed 1:1 and 2:1 Aβ₄₀: hemin complex was observed. However, at 20 and 50 equiv., the equilibrium still favored the 2:1 Aβ₄₀: hemin complex with a minor 1:1 Aβ₄₀: hemin component, as inferred from the observation of the λ_max at 414 nm signal with a reduced absorption intensity. The partial reversibility of this equilibrium was also demonstrated in the presence of 5 and 10 equivalents of Aβ₄₀ by increasing the temperature to 37 °C (ESI Figure S4).

The peroxidase activity of Aβ₄₀-hemin constructs typically reflect the coordination environment [13,14,15,16,17,18,19,20,21,22,23,24]. We thus sought to investigate how extending the pH and Aβ₄₀ concentration ranges, up to 100 equivalents, affected peroxidase activity (

Table 2. Kinetic constants for the oxidation of ABTS (λ = 660 nm; ε = 14,700 M⁻¹ cm⁻¹) [36] by H₂O₂ (2.5 mM) and hemin (2 μM) at 37 °C in the presence of varying amounts of Aβ₄₀ or Aβ₁₆ at pH 7.4 or 6.3. The calculated variability based on 2 independent measurements was equal to or less than 0.001 s⁻¹.

Species	pH 7.4−Aβ16 kobs (s−1)	pH 7.4−Aβ40 kobs (s−1)	pH 6.3−Aβ16 kobs (s−1)	pH 6.3−Aβ40 kobs (s−1)
Hemin	0.013	0.013	0.016	0.016
Aβ (1 equiv.)	0.022	0.018	0.034	0.027
Aβ (5 equiv.)	0.022	0.021	0.042	0.033
Aβ (20 equiv.)	0.032	0.027	0.053	0.038
Aβ (100 equiv.)	0.054	0.032	0.092	0.049

). With such an objective in mind, the peroxidase activity of hemin was measured by monitoring the oxidation of ABTS by hydrogen peroxide at increasing ratios of Aβ₄₀: hemin at pH 7.4 in phosphate buffer (50 mM) at 37 °C (Figure 2). The reaction rates were obtained by dividing the slope of the trace in the initial seconds of the reaction by the product of hemin concentration multiplied by the molar extinction coefficient of the ABTS radical cation multiplied by a factor of two to account for the two ABTS radical cations produced during each reaction cycle (cf. Materials and Methods). In this regime, pseudo-first order rate behavior was observed. It was concluded that at higher Aβ₄₀ concentrations, where the six-coordinate Aβ₄₀: hemin (2:1) species is reported as the predominant species [13], a greater than 2.5- and 3-fold increase in rate was observed at pH 7.4 and 6.3, respectively.

As can be seen from an inspection of Table 2 (first entry), the catalytic activity of free hemin for the oxidation of ABTS was higher at pH 6.3 (k_obs = 0.016 s⁻¹) than at 7.4 (k_obs = 0.013 s⁻¹). The addition of Aβ₄₀ to hemin further enhanced the peroxidase activity at both pH values, with k_obs always larger at pH 6.3 as compared to 7.4 (Figure 2). Previous studies detailing the oxidative reactivity of Aβ₄₀-hemin complexes have failed to explore high concentrations of Aβ₄₀ (i.e., more than 20 equivalents) [24].

Similar to the trend observed with Aβ₁₆ by Flemmig and co-workers [21], in comparison to free hemin, at 100 equiv. of Aβ₄₀, the initial rate increased by 2.5- and 3-times at pH 7.4 (0.032 s⁻¹) and 6.3 (0.049 s⁻¹), respectively. Notably, over the course of the reaction, the rate of peroxidase activity decreased, this is attributed to hemin degradation. The two conditions also showed different rates for decomposition with the larger drop in rate at pH 7.4. At pH 6.3, the peroxidase activity remained high even at longer reaction times, allowing for increased product formation (Figure 2).

Previous work by Ghosh, Dey, and co-workers has shown that Au electrode immobilized hemin-Aβ aggregates can produce partially reduced oxygen species (PROS) [37]. We thus sought to further elaborate on the peroxidase activity of aggregated Aβ-hemin at different pH. Qualitative reactivity experiments utilizing ABTS with hemin and a suspension of mixed soluble and insoluble aggregated Aβ₄₀ revealed a rate enhancement versus hemin alone (Supplementary Materials, Figure S5). A faster rate was also again seen at pH 6.3 versus pH 7.4 and when higher levels of Aβ₄₀ were present (Supplementary Materials, Figure S6). In order to explore the nature of this activity, hemin-uptake experiments, as measured by following the hemin Soret-band (λ_max = 390 nm), revealed that at pH 6.3, less hemin was absorbed within the aggregate and at a slower rate (Supplementary Materials, Figure S7). Based on these spectroscopic studies, it was inferred that at pH 6.3 some hemin also leached back into solution after the initial inclusion event.

In line with previous work by the groups of Atamna [14] and Xu [35], studies involving the use of circular dichroism (CD) spectroscopy and scanning electron microscopy (SEM) provided support for the inference that hemin can disrupt the Aβ₄₀ aggregate architecture (Figure 3). Briefly, upon addition of hemin (5 μM) to Aβ₄₀ aggregates (40 μM, 8 equivalents, aged 30 days), the negative CD minimum underwent a slight decrease in intensity with the λ_max shifting from 218 nm to 220 nm. Over the course of 24 h, a further reduction in the CD intensity was seen with a shift in the minimum to ca. 217 nm.

SEM analyses of the aggregated Aβ₄₀ (aged 30 days plus 24 h) sample showed fibrillar constructs, observable at 5 μm. The Aβ₄₀ sample (aged 30 days) incubated with hemin (5 μM) for 24 h displayed a more globular like structure, observable at 3 μm. A control experiment involving freshly dissolved Aβ₄₀ showed amorphous species, observable at 2 μm. We thus propose that the reactivity of aggregated Aβ₄₀-hemin complexes may operate under a mechanism wherein both hemin exposed on the aggregate surface and newly formed soluble oligomeric Aβ₄₀-hemin complexes contribute to the peroxidase activity. The ability of acidic pH to disrupt the β sheet architecture is also expected to enhance formation of soluble oligomeric Aβ₄₀-hemin complexes with increased peroxidase activity [37,38,39].

The monoamine neurotransmitter, dopamine (DA), plays a key role in proper neuronal function [40]. Recent studies have also demonstrated that DA and oligomeric melanin help maintain standard physiological functions and have a pathological relevance in neurodegenerative conditions [41]. We thus sought to determine how the enhanced peroxidase activity of Aβ₄₀-hemin complexes might oxidatively modify DA. The DA modifying activity of these Aβ₄₀-hemin complexes was evaluated by monitoring the formation of dopaminochrome, an intermediate during DA oxidation to melanin [42]. Notably, DA oxidation mediated by hemin and Aβ₄₀-hemin complexes did not exhibit a simple Michaelis-Menten behavior, but instead showed evidence of a substrate inhibition effect (ESI Figure S8). The kinetic constants (pseudo-first order approach) (

Table 3. Kinetic constants for the oxidation of DA (3 mM, λ = 470 nm; ε = 3300 M⁻¹ cm⁻¹) [43] by H₂O₂ (2.5 mM) and hemin (2 μM) at 37 °C with varying amounts of Aβ₄₀ or Aβ₁₆ in phosphate buffer (50 mM) at pH 7.4 or 6.3. The calculated variability, based on at least 2 independent measurements, was in all cases ≤ 0.001 s⁻¹.

Species	pH 7.4−Aβ16 kobs (s−1)	pH 7.4−Aβ40 kobs (s−1)	pH 6.3−Aβ16 kobs (s−1)	pH 6.3−Aβ40 kobs (s−1)
Hemin	0.053	0.052	0.018	0.018
Aβ (1 equiv.)	0.049	0.058	0.024	0.023
Aβ (5 equiv.)	0.056	0.060	0.023	0.027
Aβ (20 equiv.)	0.077	0.068	0.037	0.031
Aβ (100 equiv.)	0.171	0.076	0.076	0.034

) were obtained by dividing the slope of the trace in the initial seconds of the reaction by the product of hemin concentration multiplied by the molar extinction coefficient of dopamine (cf. Materials and Methods).

At pH 7.4 in phosphate buffer (50 mM), Aβ₄₀ increases the peroxidase activity of hemin in terms of DA oxidation, albeit the rate at 100 equivalents of Aβ₄₀ is only 1.5-times greater than hemin alone (Figure 4). At pH 6.3, DA oxidation becomes more difficult due to the proton-dependent increase in the substrate E° value [44]. This difference is reflected in the hemin-mediated oxidation of DA that proceeds at a 3-times slower rate of 0.018 s⁻¹ at pH 6.3 vs. 0.053 s⁻¹ at pH 7.4 (Figure 4). However, at pH 6.3, the presence of Aβ₄₀ (100 equivalents) increased the rate almost two-fold.

Similar studies utilizing the Aβ₁₆ model system showed marked differences versus Aβ₄₀. At both pH 7.4 and 6.3, Aβ₁₆ proved more effective at increasing the rate of hemin-catalyzed oxidation of ABTS and DA in the presence of H₂O₂ (Table 2 and Table 3, Figures S9–S12). In particular, Aβ₁₆ showed a significant rate enhancement with a 4-fold and almost 6-fold increase of ABTS oxidation (hemin vs. 100 equiv. Aβ₁₆) at pH 7.4 and 6.3, respectively. In the case of DA oxidation, Aβ₁₆ displayed a 3-fold and 4-fold increase at pH 7.4 and 6.3, respectively.

These findings may be rationalized by assuming mechanistically that the peroxidase-like activity of hemin resembles that of the peroxidase enzyme [45]. Briefly, there is an initial activation of H₂O₂ with attendant formation of a high valent P^•+Fe^IV=O species (Compound I), where P^•+ indicates the porphyrin π-radical cation (Equation (1)). The involvement of compound I-like oxidant for heme-Aβ peptides complexes has been recently characterized [46]. The formation of this active species is followed by two one-electron oxidations of two substrate molecules (SH) (Equations (2) and (3)). The rate-determining step (rds) of the catalytic cycle depends on the relative concentrations of H₂O₂ and SH. The experiments reported here were performed under conditions where the H₂O₂ concentration has reached kinetic saturation, making step 3 the rds. The data supports the conclusion that the rate increase seen with a large excess of Aβ₄₀ is due to faster electron transfer from SH to the Fe^IV=O species (Equation (3)).

PFe^III + H₂O₂ ⇄ [PFe^III-H₂O₂] → P^•+Fe^IV = O + H₂O

(1)

P^•+Fe^IV = O +SH → PFe^IV = O + S^• + H⁺

(2)

PFe^IV = O + SH + H⁺ → PFe^III + S^• + H₂O

(3)

The effect of Aβ peptides in promoting a marked rate increase may be ascribed to the following three hypotheses: (i) axial coordination of the Fe^III center to enhance oxidative reactivity, (ii) acid/ base catalysis facilitating the release of the oxo group as water, and (iii) enhancing substrate approach to the iron-oxo porphyrin. In order to assess the relative contributions of the previous hypotheses, analogous DA oxidation experiments were carried out using hemin-glycyl-L-histidine methyl ester (hemin-GH) [47]. The appended histidine residue is strongly bound to the iron(III) centre via an intramolecular coordination, leading to a five-coordinate iron(III) species with enhanced peroxidase activity. Not surprisingly, the rate of DA oxidation was more than one order of magnitude larger (0.785 s⁻¹) than that of free hemin under the same conditions (0.053 s⁻¹) (Supplementary Materials, Table S1). The addition of 1 or 5 equivalents Aβ₁₆ failed to modify the reaction rate significantly, with values of 0.819 s⁻¹ and 0.824 s⁻¹, respectively, being observed (Supplementary Materials, Figure S13). However, addition of 20 and 100 equivalents Aβ₁₆ caused a rate decrease, with values of 0.773 s⁻¹ and 0.591 s⁻¹, respectively, now being obtained (Table S1). This leads us to suggest that the unstructured organization of the Aβ₁₆ peptide chain around the hemin centre (hypotheses ii and iii) has little effect, if any, on the reaction rates. High levels of peptide are believed to introduce some steric hindrance, but the principal effect leading to the increase in activity is ascribed by default to iron chelation by the Aβ peptide (hypothesis i).

To the extent such considerations are correct, the mechanism of substrate oxidation by hemin in the presence of Aβ peptides must include a pre-equilibrium step leading to a mixture of hemin, hemin-Aβ, and hemin(Aβ)₂, defined by binding constants K₁ and K₂ (Supplementary Materials, Scheme S2). Upon reacting with H₂O₂, hemin is transformed into the iron-oxo species through pathway a. Hemin-Aβ and hemin(Aβ)₂ are transformed into the same active iron-oxo species through pathways b and c. The reversible disassociation of six-coordinated hemin(Aβ)₂ in the pre-equilibrium prevents reaction inhibition and ultimately yields the same active intermediate formed by hemin-Aβ (1:1). At saturating levels of H₂O₂, the distal histidine of the second Aβ peptide is also readily displaced by H₂O₂. The first Aβ peptide has been shown to coordinate the iron center strongly via one of the three histidine residues (preferably His13) [18]. This, in turn, increases the electron density at iron and favours electron transfer from the substrate to the iron-oxo species (hypothesis i).

A comparison between the two amyloid peptide species considered in the present study reveals that hemin has a higher affinity (K₁) for Aβ₄₀ over Aβ₁₆ (Table 1). However, analysis of the k_obs values reported in Table 2 and Table 3 indicate a higher activating effect (peroxidase activity) in the case of Aβ₁₆. This phenomenon is attributed to increased steric bulk around the metal center caused by the larger Aβ₄₀ peptide, something that prevents substrate approach and electron transfer. It should also be noted that a larger concentration of the Aβ is required for complete binding to hemin under conditions of catalysis than might be expected based on the value for log K₁. This is rationalized in terms of a high substrate concentration that competitively prevents binding the second hemin axial site. This, in turn, precludes catalyst activation with H₂O₂ and is also reflected in the inhibitory effect of excess DA.

Finally, we evaluated the competitive self-oxidation of Aβ in the presence of DA and H₂O₂. Previous work by our group [13,48], as well as others [49], has demonstrated that the Aβ peptide is a substrate for metal-catalyzed oxidation. Briefly, a reaction mixture of hemin (6 μM), Aβ₁₆ (30 μM), H₂O₂ (1 mM) and DA (3 mM) in phosphate buffer (50 mM, pH 7.4) was incubated at 37 °C for 30 and 120 min, followed by HPLC-MS analysis. As shown in Table S2 and Figure S14, the peptide remains mostly unmodified (79% intact) at both 30 and 120 min. Control experiments without DA showed that the modification pattern is similar, 85% unmodified Aβ₁₆, indicating that the oxidation is mainly mediated by the oxidative hemin-Aβ₁₆-H₂O₂ system (Table S3). The similar amounts of modified peptide seen for the two reaction times leads us to suggest rapid modification of Aβ₁₆ followed by catalyst decomposition or inhibition in the case of added DA.

3. Materials and Methods

3.1. General Procedures

Hemin, 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) (ABTS), and dopamine hydrochloride were purchased from Sigma Aldrich and used as received. Hemin stock solutions were prepared by dissolving hemin in 0.5 M NaOH in Milli Q water and centrifuging at 14,000 rpm for 2 min. The supernatant was decanted and diluted in Milli Q water to yield an approximately 3 mM stock solution. The phosphate buffer solution was prepared by dissolving the appropriate amounts of Na₂HPO₄ and NaH₂PO₄ then adjusting the pH with concentrated NaOH (aq.). Aβ₁₆ was prepared as previously reported [48]. Aβ₄₀ was prepared via Fmoc amino solid-phase peptide synthesis using a Liberty Blue™ microwave peptide synthesizer [50]. Preparative RP-HPLC purification of peptides was performed using an Agilent Zorbax SB-C₁₈ Prep HT column 21.2 × 250 mm. Analytical RP-HPLC characterization of peptides was performed using an Agilent Zorbax column 4.6 × 250 mm. An Agilent Technologies 6530 Accurate Mass QT of LC/MS was used for high-resolution mass spectra of purified peptides. Solvents used were HPLC grade. Aβ₄₀ stock solutions for UV-vis titrations were prepared by dissolving the peptide in Milli Q water. The solution was sonicated and vortex mixed until the peptide was dissolved. This mixture was filtered through a 0.22 micron PTFE filter. Concentration of the Aβ₄₀ peptide was measured by UV-vis spectroscopy using a molar extinction coefficient ε₂₈₀ = 1480 M⁻¹ cm⁻¹ [51]. Concentration of hemin solution was measured by UV-vis spectroscopy using a molar extinction coefficient ε₃₉₀ = 66,000 M⁻¹ cm⁻¹ [52]. UV-vis spectra and kinetic experiments were recorded on a Hewlett Packard HP 8453A diode array spectrophotometer equipped with a thermostated, magnetically stirred cuvette holder. The peptides were studied by direct LC-MS/MS analysis. LC-MS and LC-MS/MS data were obtained by using an LCQ DECA ion-trap mass spectrometer with an ESI ion source, coupled with an automatic injector Surveyor HPLC system and controlled by the Xcalibur 1.3 software (Thermo-Finnigan, San Jose, CA, USA). The system was run in an automated LC-MS/MS mode and by using a Surveyor HPLC system (Thermo-Finnigan, San Jose, CA, USA) equipped with a BIOBASIC C18 column (5 µM), 150 × 2.1 mm). For analysis of oxidation studies, the acquired MS/MS spectra were automatically searched against the human Aβ₁₆ sequence by using the SEQUEST algorithm to identify the modified residues. This algorithm was incorporated into the Bioworks 3.1 software (Thermo-Finnigan, San Jose, CA, USA). SEM analyses were prepared by drop-casting samples onto silica wafers (University Wafers) then gently drying with nitrogen. Images were taken on a FEI Quanta 650 ESEM instrument in The University of Texas at Austin–Texas Materials Institute. Circular dichroism (CD) analyses were carried out using a Jasco J-815 CD spectrometer acquired through The University of Texas at Austin Texas Institute for Drug Discovery & Development (Ti3D). Titration data were processed with the Hyperquad package [53] using a range of wavelengths centered around the two absorption maxima. The fittings allowed the determination of the equilibrium constants for the two-step binding processes:

Step 1:   [hemin(H2O)2] + Aβ40 ⇄ [hemin(Aβ40)(H2O)] + H2O  (K1)

Step 2:   [hemin(Aβ40)(H2O)] + Aβ40 ⇄ [hemin(Aβ40)2] + H2O  (K2)

Global:   [hemin(H2O)2] + 2 Aβ40 ⇄ [hemin(Aβ40)2] + 2 H2O  (β2)

3.2. Kinetics of ABTS Oxidation

The catalytic oxidation of ABTS by a hemin-hydrogen peroxide mixture was studied at 37 °C in 50 mM phosphate solution at pH 7.4 or 6.3. ABTS (3 mm) oxidation was initiated by the addition of hemin (3 µM), H₂O₂ (2.5 mM) and amyloid-β fragments 1–16 and 1–40 (0–200 µM). Each mixture containing hemin and buffer solution was equilibrated at 37 °C for 5 min before the addition of the substrate and peptide; hydrogen peroxide was added to the reaction solution as the last reagent. The reactions were followed through the development of the optical band of ABTS^+· radical cation at 660 nm (ε₆₆₀ = 14,700 M⁻¹ cm⁻¹) [36]. Reaction rates in Δ(absorbance/s) units were calculated from the slope of the trace in the initial 30 s, with the first few seconds discarded to allow stabilization of the readings, and were then converted into the observed kinetic constant (s⁻¹) by dividing the calculated value over the product of catalyst concentration multiplied by the molar extinction coefficient of ABTS^·+ multiplied by a factor of two (each catalytic cycle involves formation of two ABTS^·+ molecules).

3.3. Kinetics of Dopamine (DA) Oxidation

The catalytic oxidation of DA by a hemin-hydrogen peroxide mixture was studied at 37 °C in 50 mM phosphate solution at pH 7.4 or 6.3. DA (3 mm) oxidation was initiated by the addition of hemin (3 µM), H₂O₂ (2.5 mM) and amyloid-β fragments 1–16 and 1–40 (0–200 µM). Each mixture containing hemin and buffer solution was equilibrated at 37 °C for 5 min before the addition of the substrate and peptide; hydrogen peroxide was added to the reaction solution as the last reagent. The reactions were followed via the development of the optical band of dopaminochrome at 470 nm (ε₄₇₀ = 3300 M⁻¹ cm⁻¹) [43]. Reaction rates in Δ(absorbance/s) units were calculated from the slope of the trace in the initial 30 s, with the first few seconds discarded to allow stabilization of the readings, and were then converted into the observed kinetic constant (s⁻¹) by dividing by the product of the catalyst concentration multiplied by the molar extinction coefficient of dopaminochrome.

3.4. Aggregate Formation

The aggregated form of Aβ₄₀ was prepared as follows: a stock solution of Aβ₄₀ was prepared by dissolving Aβ₄₀ in DMSO (ca. 700 µM). The mixture was sonicated and vortex mixed until the peptide was solubilized. This solution was diluted with 20 mM phosphate buffer solution (pH 7.5), to give a final concentration of <2% DMSO, and incubated at 37 °C for 7 days. The formation of insoluble aggregate species was observed, as seen by the appearance of insoluble Aβ₄₀ species. The mixed soluble and insoluble Aβ₄₀ aggregate solution was split into Eppendorf tubes and centrifuged down. The supernatant was decanted and the solid was washed three times with Milli Q water (1 mL) to remove any salt and DMSO. The aggregate was resuspended in a UV-vis vial and utilized for the peroxidase activity studies. Two sample concentrations were prepared by varying the level of dilution.

3.5. Aβ₁₆ Modification in the presence of Hydrogen Peroxide and Hemin by LC-MS

Competitive peptide modification was studied using HPLC-ESI/MS. Samples were prepared by mixing hemin (6 μM), H₂O₂ (1 mM), Aβ₁₆ (30 μM), and DA (3 mM) in phosphate buffer solution (50 mM) pH 7.4 and incubating at 37 °C. HPLC-ESI/MS analysis were performed at different reaction times. LC-MS data were obtained by using a LCQ ADV MAX ion-trap mass spectrometer. The elution of Aβ₁₆ was carried out by using 0.1% HCOOH in distilled water (solvent A) and 0.1% HCOOH in acetonitrile (solvent B), with a flow rate of 0.2 mL/ min. The solvent gradient started with 98% solvent A for 5 min followed by a linear gradient from 98 to 55% A over 65 min.

4. Conclusions

In conclusion, we have demonstrated that the amyloid peptide Aβ₄₀ displays an environmental dependence on the dynamic equilibrium between high-spin five-coordinate hemin-Aβ₄₀ and low-spin six-coordinate hemin(Aβ₄₀)₂ species and that this affects the peroxidase activity of these species. Hemin was also shown to disrupt Aβ₄₀ aggregated structures and display enhanced peroxidase activity in the presence of Aβ₄₀ aggregates versus hemin alone. This work thus sets the stage for understanding in greater detail the role of aberrant metal-ions (e.g., Fe, Cu, and Zn) in relationship to both full length and smaller hydrolyzed Aβ peptides found in Alzheimer’s disease brains and the role of the associated complexes in enhancing reactive oxygen (ROS) and nitrogen species (RNS) that are considered causal in CNS neurodegeneration [54]. Furthermore, within the context of this and related studies, the enhanced peroxidase activity of hemin-peptide constructs are counterproductive to normal healthy activity, wherein cells maintain an array of corrective mechanisms to prevent cellular and organismal death. New studies designed to understand the specific processes resulting in these phenomena (i.e., the release of amyloid beta peptides and dysregulation of hemin pathways resulting in hemin-amyloid beta species) may uncover key processes related to the origins of Alzheimer’s and related neurodegenerative diseases.