Metal Binding by GMP-1 and Its Pyrimido [1,2]benzimidazole Analogs Confirms Protection Against Amyloid-β Associated Neurotoxicity

Abstract

Alzheimer’s disease (AD) represents a major public health threat and, unfortunately, available therapeutics provide only temporary symptomatic relief. AD is a complex multifactorial disease and failure of single target therapeutics targeting amyloid-β (Aβ) in recent clinical trials suggests that future AD drug development should be focused on simultaneous targeting of several pathological hallmarks of the disease. Recently, we have shown that GMP-1, a 2-(methoxymethyl)pyrimido [1, 2-a] benzimidazol-4-ol, protects mitochondrial function in drosophila and mice models of AD, and improved memory and behavior indicating neuroprotective effect of GMP-1 treatment. Here, we have found that GMP-1 specifically binds to copper and zinc, metals that are dysregulated in AD brain. Addition of GMP-1 does not inhibit metal-dependent enzymatic reactions. Also, binding of Zn(II) and Cu(II) by GMP-1 is weaker than the 8-hydroxyquinoline scaffold compound clioquinol previously tested in AD clinical trials. However, GMP-1 affects Cu(II)-dependent Aβ fibrillization as well as oxidative damage and viability of SH-SY5Y cells upon addition of Cu(II) and Aβ. Our data provide new insight on GMP-1 as a Zn(II) and Cu(II) specific metal chelator of moderate affinity that can be responsible for some of its neuroprotective effects observed in AD animal models.

Keywords

Alzheimer’s disease metal chelation multi-target-directed ligands treatment strategies

INTRODUCTION

Alzheimer’s disease (AD) is a major form of dementia with a worldwide prevalence of more than 50 million people which is expected to rise to 152 million in 2050 [1]. AD presents an enormous unmet medical need and unfortunately, no effective treatments because of its complex and poorly understood pathogenesis and limited effectiveness of single target treatment approaches. The etiology of AD is multifactorial and its pathophysiology is complex, involving disturbances and imbalances in various mechanisms, which ultimately lead to neuronal death. The two main pathological hallmarks of AD are senile plaques deposited around neurons and neurofibrillary tangles, which are twisted fibers inside neurons [2] associated with protein misfolding, thereby leading to the extracellular aggregation of amyloid-β (Aβ) peptides and intracellular aggregates of tau protein, respectively. However, other major pathological conditions also play important roles in the progress of the disease, including enhanced brain oxidative stress and disruption of metal homeostasis [3], glucose utilization deficiency and mitochondrial dysfunction [4, 5], as well as increased neuroinflammation and microglial activation [6]. It is becoming increasingly clear that the complex pathology of AD may be responsible for the lack of an effective pharmacological treatment. Therefore, to develop the next generation of therapies, the focus has shifted toward searching for molecules that exhibit multi-target properties to combat the multifactorial nature of the disease.

The metal hypothesis based on the therapeutic chelation approach to manage AD, hypothesizes that small organic metal chelators would hamper the metal-induced amyloid deposition and oxidative stress in the diseased brain [7]. Compelling evidence suggests critical role of copper and zinc ions in both precipitating and potentiating AD [7]. Altered metal homeostasis contributes to the loss of neurons from a complex interplay of factors including oxidative injury, excitotoxic stimulation, dysfunction of critical proteins, and mitochondrial failure. Several attempts have been made to modulate metal homeostasis in AD using small molecule metal chelators to rescue the metal-dependent Aβ aggregation. The most studied class of compounds contains 8-hydroxyquinoline (HQ) scaffold and these compounds have shown potential for the treatment of AD and other neurodegenerative disorders. Two of these, clioquinol (5-chloro-7-iodo-8-hydroxyquinoline, CQ) and PBT2 (5,7-dichloro-2-[(dimethylamino)methyl]quinolin-8-ol), have shown promising results in several clinical trials for AD but further discontinued due to their side effects like sub-acute myelo-optic neuropathy in case of CQ [8] or due to the lack of cognitive improvement for PBT2 [9]. The possible explanation behind these observations is that 8-hydroxyquinoline analogues are able to chelate virtually all transition metals with high affinity, leading to depletion of essential trace metals. For example, the mechanism of toxicity of CQ was attributed to the depletion of vitamin B₁₂ (cobalamin) [10]. Therefore, metal chelators with a moderate chelation dynamics are required to develop therapeutics for the management of transition metal-mediated amyloidosis in AD.

Recently, we have developed GMP-1, a 2-(methoxymethyl)pyrimido [1, 2-a] benzimidazol-4-ol, a compound inhibiting interaction between molecular chaperones Hsp70/Hsp90 and protein import receptor Tom70, as a potential lead compound for development of drugs against AD [11]. GMP-1 treatment of SH-SY5Y cells resulted in decreased mitochondria-associated AβPP. Further, experiments in drosophila and mice models of AD demonstrated increased fly viability, improved memory and behavior in mice as well as restored mitochondrial function indicating neuroprotective effect of GMP-1 treatment.

In this paper, we have investigated metal binding capacity of GMP-1 and other related analogs and show importance of metal chelation for neuroprotective effect of GMP-1 in Aβ_1–42 treatment paradigm.

METHODS

Chemicals

GMP-1 and example compounds were obtained from Vitas-M Laboratory, Apeldoom, The Netherlands. Pyrimido [1, 2-a]benzimidazol-4(1H)-one was synthesized at AKos GmbH, Steinen, Germany. All common chemicals were purchased from Sigma (St. Louis, MO, USA) unless stated otherwise. GMP-1 and its analogs were dissolved in DMSO and further used from this stock solution.

Determination of the metal to GMP-1 binding and stoichiometry by UV/Vis spectroscopy

UV/Vis spectra of GMP-1 and CQ were recorded with NanoDrop ND-100 spectrophotometer (Thermo Fisher Scientific). The ligand/metal stoichiometry of the GMP-1 complexes were investigated by the addition of increasing molar fractions of metal ions (Cu²⁺ or Zn²⁺) 0.2 to 2 molar equivalents of metal ion. Aliquots of aqueous solutions of CuCl₂ or ZnSO₄ were added to a stirred solution of 0.1 mM of GMP-1 and its analogues and/or 0.1 mM CQ in 0.1 M HEPES-KOH buffer, pH 7.5. pH-dependent metal binding was measured within pH range of 5.0–8.0 in 0.1 M HEPES-KOH buffer. Mg²⁺, Ca²⁺, Al³⁺, Ni²⁺, Co²⁺, Fe²⁺, Fe³⁺, Mn²⁺, and Pb²⁺ were added from stock solutions of their respective salts. The total variation of the volume in the reaction did not exceed 2%. A UV/Vis spectrum was recorded after each addition of metal salt. Changes in the spectrum occurred immediately (fast complexation) and were stable between two consecutive additions.

Limited proteolysis of β-casein

Proteolysis of β-casein with collagenase was performed as following: 0.02 mg/ml of collagenase was incubated with 0.1 % DMSO or with 50 μM of EDTA, 50 μM of GMP-1 or 50 μM of CQ for 30 min at 4°C. β-casein was added from stock solution of 10 mg/ml to final concentration of 1 mg/ml and reaction mixture was incubated for 10 min at 4°C. 2X SDS sample buffer was added to the reaction, mixture was immediately boiled for 5 min and loaded on SDS-PAGE. Gels were subsequently stained with Coumassie Brilliant Blue R-250 and photographed.

Cytochrome oxidase activity assay

Cytochrome c (2.7 mg/ml) was reduced by incubation with DTT and excess of reducing agent was removed by gel filtration on PD-10 column (GE-Healthcare, Uppsala, Sweden). Cytochrome oxidase activity was measured by decrease in absorbance of ferrocytochrome c at 550 nm. Mouse brain mitochondria were isolated according to [12]. Isolated mitochondria were diluted in the buffer containing 10 mM Tris-HCl, pH 7.0, 0.5 M sucrose, 0.05 % Triton X-100 to concentration 0.5 mg/ml and were incubated with DMSO only, 50 mM of CQ or 50 mM of GMP-1 for 10 min at 25°C. After addition of ferrocytochrome c, latency in 550 nm absorbance were immediately measured. Experiments were performed in triplicate.

Cellular viability assays

SH-SY5Y human neuroblastoma cells were obtained from the American Tissue Culture Collection (ATCC) and maintained in DMEM supplemented with 10% FBS and 1% penicillin-streptomycin (Gibco/Invitrogen, Carlsbad, CA, USA). SH-SY5Y cells were pre-differentiated with 10 μM Retinoic acid in complete media for 4 days and then treated with 50 ng/mL BDNF in serum free media for 2 days. Differentiated SH-SY5Y cells were cultured in 5% CO₂ -95% air at 37°C. Aβ₁_-₄₂ stock (50 μM) was mixed with equimolar amounts of Cu(II) in DMEM without FBS and incubated for 3 h prior to addition to the cells. Aβ_1–42 - Cu(II) complex (10 μM final conc.) was used alone or in combination with CQ (10 μM final conc.), GMP-1 (20 μM final conc.), and GMP-1 analogue (GA, 20 μM final conc.) that was unable to bind to Cu(II) and Zn(II) and were added from 100X DMSO stock solution directly to the culture media and incubated overnight. Cellular toxicity was assessed with MTT Cell Proliferation Kit I (Roche Applied Science, Indianapolis, IN, USA) according to the manufacturer’s protocol. TUNEL-DAPI co-staining assay was carried out by the in situ cell death detection kit (Fluorescein), Roche Applied Science, Indianapolis, IN, USA according to the manufacturer's protocol. Images were captured using fluorescent microscope (magnification 200X).

Reactive oxygen species (ROS) detection using DCFDA

The intracellular ROS generation was monitored by DCFDA (2’,7’–dichlorofluorescin diacetate) assay. Undifferentiated SH-SY5Y cells (1×10⁵ per well) were treated as described above in the MTT assay. Then, cells were incubated with DCFDA (10 μM) at 37°C for 30 min. The intracellular ROS formation was examined under fluorescent microscope (magnification 200X) with the excitation and emission wavelengths at 488 nm and 525 nm, respectively. The intracellular ROS levels were also analyzed by flow cytometry using Accuri^TM C6 plus flow cytometer from BD Biosciences. Cells (1 x 10⁵) were detached, washed once with PBS and subjected to flow cytometry with fluorescein filter.

Aβ_1–42 fibrillization assay

Aβ_1–42 fibril formation kinetics was studied using thioflavin T (ThT) fluorescence measurement (λex = 440 nm,λem = 490 nm) in a plate reader (FLUOStar Galaxy from BMG Labtech, Offenberg, Germany). The fluorescence was recorded using bottom optics in half-area 96-well polyethylene glycol-coated black polystyrene plates with clear bottom (Corning Glass, 3881). Aβ_1–42 monomer was isolated by size exclusion chromatography over a Superdex 75 column (GE Healthcare) in 20 mM sodium phosphate, 0.02% NaN₃ at pH 8 and kept on ice. Every sample was supplemented with 10 μM ThT from a 1 mM stock solution. 3 μM of Aβ_1–42 was incubated with 3 μM of CuCl₂ for 4 h at 25°C followed by addition of 0.01 % of DMSO alone or 3 μM of GMP-1. Fluorescence was measured every 5 min during 16 h. Standard deviation of four measurements is shown.

RESULTS

Metal binding properties of GMP-1

Several transition metal binding compounds utilizing 8-hydroxyquinoline scaffold including CQ and PBT2 have recently been reported as potential drug candidates for the treatment of AD and other neurodegenerative diseases [13]. Structural studies revealed that metals are coordinated with nitrogen and oxygen atoms of CQ [14]. We have found considerable structural similarities between CQ and GMP-1 (Fig. 1A) prompting us to investigate metal binding properties of GMP-1 and its pyrimido [1, 2]benzimidazole analogs in better detail. Figures 1B and 1C represent absorbance spectra of 0.1 mM GMP-1 in presence of increasing amounts of Zn(II) and Cu(II), respectively. GMP-1 has two absorbance maximums at 230 nm and 330 nm. Addition of increasing amounts of Zn(II) and Cu(II) caused decrease in absorbance at 230 nm and 330 nm. The presence of several isobestic points at 260 nm, 275 nm, and 340 nm revealed formation of Zn(II) GMP-1 complex consistent with 1 : 1 metal ligand ratio (Fig. 1B). Similar pattern was observed upon titration of GMP-1 solution with Cu(II) until the 1 : 1 metal ligand ratio; however, addition of higher concentrations of Cu(II) resulted in shift of spectra maximums indicating formation of higher order complexes between Cu(II) and GMP-1 (Fig. 1C). Next, we have investigated ability of GMP-1 to bind to other transition metals as well as magnesium and calcium ions. Unlike CQ, which showed change in UV/Vis absorption spectrum in the presence of metals, none of the metals including Mg²⁺, Ca²⁺, Al³⁺, Ni²⁺, Co²⁺, Fe²⁺, Fe³⁺, Mn²⁺, and Pb²⁺ elicit any changes in UV/Vis absorbance spectrum of GMP-1 (data for Ni²⁺, Co²⁺, and Mn²⁺ are shown in Fig. 1D) indicating binding selectivity of GMP-1 toward Zn(II) and Cu(II). GMP-1 binding of Zn(II) and Cu(II) occurs at pH above 6.5 (Fig. 1E) suggesting importance of deprotonated state of exocyclic oxygen for efficient metal binding. We have subsequently investigated 45 structural analogues of GMP-1 that possess common pyrimido [1, 2]benzimidazole scaffold for their ability to bind Zn(II) and Cu(II) (Supplementary Table 1). Structure-activity analysis of pyrimido [1, 2] benzimidazole analogues of GMP-1 revealed that the substitution pattern affects the metal chelating capacity to a great extent.

Fig.1

A) Comparison of structures of CQ with GMP-1 and related compounds. The oxygen and nitrogen metal coordination sites for CQ are shown as dashed lines. Potential metal coordination site for GMP-1 and analogues is shown by dashed circle. B) GMP-1 UV-vis titration with Zn(II). The 0.1 mM GMP-1 solution was titrated with 0.2 to 2 equivalents of ZnSO₄. Arrowhead indicates GMP-1 spectrum without Zn(II), arrow indicates GMP-1 spectrum upon addition of 0.2 mM of ZnSO₄. C) GMP-1 UV-vis titration with Cu(II). The 0.1 mM GMP-1 solution was titrated with 0.2 to 2 equivalents of CuCl₂. Arrowhead indicates GMP-1 spectrum without Cu(II), arrow indicates GMP-1 spectrum upon addition of 0.2 mM of CuCl₂. D) GMP-1 (0.1 mM) UV-vis spectrum in absence (black), and in presence of equimolar conc. of Mn²⁺ (red), Ni²⁺ (green) and Co²⁺ (orange) ions. E) pH-dependent interaction of GMP-1 with Cu/Zn. The absorbance of 0.1 mM GMP-1 solution at λ= 330 nM, pH 8.0 was set to 100 %.

Effect of GMP-1 versus metal chelators on metal-dependent enzymatic activities

High affinity metal chelating compounds can potentially have detrimental effect on living organisms removing essential trace metals from the enzymes and body fluids. CQ, a small lipophilic molecule, possesses moderate affinity for Cu(II) and Zn(II) (K_d of 8×10^–11 M and 1.4×10^–9 M, respectively) [15]. We have probed Cu(II) and Zn(II) binding to GMP-1 in the presence of equimolar amounts of CQ. UV/Vis absorbance spectra measurement revealed no changes in absorbance spectrum of GMP-1 in the presence of both metals and CQ (data not shown) indicating that all metal ions were complexed with CQ. To investigate the effect of GMP-1 and other metal chelators on metal-dependent enzymatic reactions, we performed two separate assays; limited proteolysis using Zn(II)-dependent collagenase and mitochondrial Cu(II)-dependent cytochrome c oxidase activity assay. Cleavage of β-casein by collagenase was performed in the absence or in the presence of metal chelators EDTA, CQ, or GMP-1 (Fig. 2A). Unlike EDTA and CQ that inhibit collagenase activity, GMP-1 did not affect formation of β-casein fragment marked by small arrow (Fig. 2A), indicating inability of GMP-1 to deplete Zn(II) from active site of collagenase. We further investigated the ability of GMP-1 to chelate Cu(II) from mitochondrial cytochrome oxidase, measuring copper-dependent cytochrome oxidase activity. Addition of CQ inhibited cytochrome oxidase activity by 25 % whereas GMP-1 did not affect it (Fig. 2B). Collectively, these data suggest low affinity complex formation of GMP-1 with copper and zinc ions.

Fig.2

A) SDS-PAGE results of β-casein cleavage with collagenase in the absence or presence of various metal chelators. Lane 1: β-casein alone, lanes 2–5 β-casein cleavage with collagenase, lane 2: no chelator addition, lane 3: collagenase preincubated with 50 μM EDTA for 10 min, lane 4: collagenase preincubated with 50 μM GMP-1 in DMSO for 10 min, lane 5: collagenase preincubated with 50 μM CQ for 10 min. Final concentration of DMSO was 0.1%. B) Cytochrome oxidase activity measurement of isolated mouse brain mitochondria incubated with DMSO only or with 50 μM CQ and 50 μM of GMP-1. Cytochrome oxidase activity in presence of 0.1% DMSO was referred as 100%. One-way ANOVA followed by Tukey’s multiple comparisons test was used (n = 3 ^*p < 0.001).

Effect of GMP-1 on Aβ₁_-₄₂ copper binding and copper-dependent Aβ₁_-₄₂ aggregation

There is considerable controversy regarding the affinity of Cu(II) and Zn(II) to Aβ_1–40 and Aβ_1–42 peptides depending on the technique applied to determine this value and the type and the concentration of the buffer used [16]. A direct determination of the dissociation constant for the Cu(II) complex with Aβ_1–40 peptide produced $K_{d}^{cond}$ = 57±5 nM [17], whereas the dissociation constant K_d = 107 nM for the Zn(II) complex with Aβ_1–40 peptide was reported [18]. This indicate at least two orders of magnitude difference in metal binding between CQ and Aβ where the latter represents a weaker metal chelator. We have compared affinities of GMP-1 and Aβ_1–42 toward Cu(II) (Fig. 3A). First, we have measured UV-vis absorbance spectrum of GMP-1 alone, GMP-1 in the presence of equimolar Cu(II) or GMP-1 in the presence of equimolar Cu(II) and Aβ_1–42. Aβ_1–42 alone does not absorb light at λ= 330 nm (Fig. 3A). Addition of Cu(II) decreased GMP-1 absorbance at 330 nm whereas presence of equimolar Aβ_1–42 increased absorbance at 330 nm by 60% (Fig. 3A). These results indicate that GMP-1 and Aβ_1–42 have similar affinities for the Cu(II). Next, we have investigated effect of GMP-1 on Cu(II)-dependent Aβ_1–42 fibrillization. It is well known that Cu(II) prevents Aβ from adopting a β-sheet conformation [19 –21] and promotes formation of more toxic oligomeric Aβ aggregates [19]. Presence of Cu(II) inhibited Aβ_1–42–dependent thioflavin T (ThT) fluorescence whereas addition of equimolar GMP-1 partially restored ThT signal (Fig. 3B). This indicates ability of GMP-1 to interfere with Aβ_1–42–Cu(II) interactions.

Fig.3

GMP-1 interferes with Aβ_1–42 complexation of Cu(II) and reduces effect of Cu(II) on Aβ_1–42 fibrillization kinetics. A) UV-vis absorbance measurement of Aβ_1–42 alone (blue), GMP-1 alone (red), GMP-1 in presence of CuCl₂ (green), and mixture of Aβ_1–42 GMP-1 and CuCl₂ (yellow). Final concentration of all components was 100 μM. B) Aggregation kinetics as monitored by ThT fluorescence of 3 μM Aβ₄₂ in the absence (green) or presence of 3 μM of Cu(II) (blue) or 3 μM of Cu(II) plus 3 μM of GMP-1 (red). Every kinetic trace is the average of four measurements with the standard deviations plotted as error bars on each time point.

Prevention of Aβ_1-42–Cu(II) complexes-induced ROS generation by GMP-1 and CQ

In the AD brain, Aβ aggregates activate microglia and induce ROS formation [22]. Moreover, Aβ can also react with redox-active metal ions to produce ROS [23]. ROS such as hydrogen peroxide (H₂O₂) are capable of permeating across cell membranes causing oxidative injury associated with AD. To clarify the antioxidant effects of GMP-1 in conditions associated with Aβ_1–42–Cu(II) complexes-induced ROS formation, we have measured the accumulation of ROS in differentiated SH-SY5Y human neuroblastoma cells by DCFDA assay (Fig. 4). While Cu(II) alone did not induced ROS production in SH-SY5Y cells (not shown), the combined treatment of cells with Aβ_1–42 and Cu(II) increased ROS formation by twofold. GMP-1 and CQ addition significantly reduced fluorescence signal in the cells. To investigate whether the metal complexation by GMP-1 mediates its inhibitory effect on Aβ_1–42–Cu(II) induced ROS formation, we used 2-(trifluoromethyl) pyrimido [1, 2-a] benzimidazol-(1H)-4one, GMP-1 analogue (GA) that does not bind to Zn(II) or Cu(II) (see Supplementary Table 1). Addition of GA did not reduce fluorescence signal in the treated cells suggesting that inhibition of ROS formation by GMP-1 depends on its metal binding capacity.

Fig.4

A) Detection of ROS formation by fluorescent imaging of SH-SY5Y cells treated with Aβ_1–42 alone or in combination with Cu(II), GMP-1, CQ, and GMP-1 analogue (GA) using DCFDA. B) Quantitative analysis of ROS levels in SH-SY5Y cells treated with Aβ_1–42 alone or in combination with Cu(II), GMP-1, CQ, and GMP-1 analogue (GA) by flow cytometry. One-way ANOVA followed by Tukey’s multiple comparisons test was used (n = 3 ^*p < 0.001).

GMP-1 and CQ inhibit SH-SY5Y neuroblastoma cell death treated with Aβ_1–42–Cu(II) complexes

It has been previously reported that Cu(II) increases Aβ_1–42-induced cell toxicity [24, 25]. We have investigated ability of GMP-1, GA, and CQ to protect cells against apoptosis induced by the treatment with Aβ_1–42–Cu(II) complex. We have applied TUNEL-DAPI co-staining to observe nuclear apoptosis. DNA fragmentation in the early stages of cell apoptosis can be stained in green by TUNEL. As indicated in Figure 5A, Aβ_1-42–Cu(II) complex significantly increased the number of nuclei stained in green. This suggests that Aβ₄₂ aggregates and Aβ₄₂–Cu(II) complex are capable of inducing early apoptotic death in SH-SY5Y cells. In the presence of GMP-1 and CQ a significant decrease in DNA fragmentation was observed, whereas GMP-1 analogue GA was unable to decrease the amount of nuclear apoptosis. We have obtained similar results using MTT cell viability assay (Fig. 5B). Formazan staining was decreased by 60 % upon treatment with Aβ_1–42–Cu(II) complexes. GMP-1 and CQ, but not GA, could reduce the detrimental effect of Aβ_1–42–Cu(II) complex on cell viability. In both cell death assay and MTT cell viability assay, there was no difference between control and metal only groups (data not shown). Taken together, our results suggest that part of GMP-1 biological activities in AD-related conditions can be attributed to the complexation of transition metals such as Zn(II) and Cu(II) ameliorating negative effects of metal imbalance in AD.

Fig.5

GMP-1 and CQ prevent Aβ_1–42–Cu(II) complexes-induced cell death. A) SH-SY5Y cells apoptosis detected by TUNEL-DAPI co-stain assay. Cell nuclei are in blue. DNA breaks indicating apoptosis are in green. B) MTT viability assay of SH-SY5Y cells treated with Aβ_1–42 alone or in combination with Cu(II), GMP-1, CQ, and GMP-1 analogue. Data represents percent of formazan absorbance in the treated samples in comparison to the control untreated sample. One-way ANOVA followed by Tukey’s multiple comparisons test was used (n = 3 ^*p < 0.001).

DISCUSSION

AD is becoming a global burden with ever-increasing aged population and absence of disease modifying therapeutics. The demand of new therapeutics for AD is greater than ever especially after continuous failure of several pharmacotherapies in the late stage clinical trials.

The “metal hypothesis” is one of the theories of AD pathogenesis largely based on the disturbed homeostasis of metals in AD patients [26]. Metal chelators have been studied extensively with the aim to restore the metal imbalance by sequestering metal ions and thereby inhibiting them to bind to the pathological proteins such as Aβ and tau to reverse the disease. CQ is among one of the most studied copper and zinc chelators which has been evaluated in the clinical trials. However, due to its side effects, it was discontinued and could not enter the clinics. The reason for its toxicity was non-selective chelation of metal ions and thereby causing alarming side effects. These observations encouraged us to discover novel chemotypes with selective bivalent metal chelating ability so as to avoid the side effects observed with previous compounds such as CQ and PBT2.

In this paper, we report GMP-1 and its analogues as a metal chelator specific for copper and zinc metal ions. GMP-1 selectively chelates the copper and zinc ions and did not show any complexation with other metal ions such as Mg²⁺, Ca²⁺, Al³⁺, Ni²⁺, Co²⁺, Fe²⁺, Fe³⁺, Mn²⁺, or Pb² ⁺ . Further, evaluation of GMP-1 analogs revealed that the metal binding capacity is affected by the substitutions on the scaffold. It has also been observed in the case of HQ derivatives where marked difference in metal chelating activity has been observed between different substituted derivatives of HQ [27].

Since metal ions also acts as cofactors for several enzymes termed as metalloproteins, it is important to have mild metal chelating activity for a candidate drug so that it does not interfere with the normal functioning of these enzymes. In a limited proteolysis assay using Zn(II) dependent collagenase and mitochondrial Cu(II) dependent cytochrome c oxidase activity assay, we found that GMP-1 and its analogs did not affect the activity of these metalloenzymes unlike the CQ and EDTA which affected the enzyme activity to a larger extent. It suggests that the GMP-1 and some of its analogs are mild metal chelators and did not produce unwanted inhibition of the other protein functions.

Moreover, GMP-1 exhibited antioxidant effect upon treatment of differentiated SH-SY5Y neuroblastoma cells with Aβ_1–42–Cu(II) complex. This effect is attributed to the metal complexation by GMP-1 since close analog of GMP-1 that did not form complex with Cu(II) or Zn(II) was unable to reduce ROS formation. We have previously identified GMP-1 as a molecule protecting mitochondrial function in AD conditions [11]. We have found neuroprotective effect of GMP-1 in drosophila and mouse AD models. In transgenic drosophila, expressing Aβ_1–42 in the brain, addition of GMP-1 improved locomotion and viability of flies [11]. It has been previously reported that toxicity of Aβ in drosophila AD models is mediated by copper and zinc [28 –30] and food supplemented with metal-chelating substances suppressed these phenotypes [30]. Therefore, neuroprotective effects of GMP-1 in drosophila models of AD can be explained, at least in part, by the GMP-1 metal complexation.

Overall, our data suggest that GMP-1 exhibit multi-target mechanism of action. Currently, no effective treatment for AD is available and this can be due to multiple factors involved in AD pathophysiology and severity. For the rational design of drug candidates, a new strategy called multi-target-directed ligands has been used to develop a variety of hybrid compounds capable to act simultaneously in diverse biological targets [31]. This includes development of bi-functional molecules and peptides with metal chelating capacity and inhibiting Aβ aggregation [32 –35], compounds inhibiting ROS production and Aβ aggregation [36], inhibitors of oxidative stress and acetylcholinesterase (AChE) [37], cholinesterases and monoamino oxidase inhibitors [38], AChE/GSK-3β inhibitors [39], and GSK-3β/BACE-1 inhibitors [40, 41]. Many of these compounds exhibit strong effect towards desired targets, low toxicity as well as favorable pharmacokinetic properties providing strong basis for the upcoming clinical trials.

In conclusion, we have reported the high specificity of GMP-1 toward chelating Zn(II) and Cu(II) metal ions avoiding indiscriminate binding to other important metal ions. In addition, GMP-1 is also capable of interfering with the Aβ_1–42–Cu(II) complex formation and has shown a protective effect in the cells against oxidative stress and apoptosis.

Footnotes

ACKNOWLEDGMENTS

This study was supported by grants from Swedish Research Council (2018–02843), Brain foundation (Fo 2017–0150), Foundation for Geriatric Diseases at Karolinska Institutet, Gunvor and Josef Anérs Foundation, Magnus Bergvalls Foundation, Gun and Bertil Stohnes Foundation, Tore Nilssons Foundation for medical research, Margaretha af Ugglas foundation, and the Foundation for Old Servants.

We are grateful to Henrik Biverstål for the help in ThT fluorescence measurements.

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

WHO, Dementia fact sheet, https://www.int/en/news-room/fact-sheets/detail/dementia, Accessed 9 June, 2019.

Winblad

, Amouyel

, Andrieu

, Ballard

, Brayne

, Brodaty

, Cedazo-Minguez

, Dubois

, Edvardsson

, Feldman

, Fratiglioni

, Frisoni

, Gauthier

, Georges

, Graff

, Iqbal

, Jessen

, Johansson

, Jonsson

, Kivipelto

, Knapp

, Mangialasche

, Melis

, Nordberg

, Rikkert

, Qiu

, Sakmar

, Scheltens

, Schneider

, Sperling

, Tjernberg

, Waldemar

, Wimo

, Zetterberg

(2016) Defeating Alzheimer’s disease and other dementias: a priority for European science and society. Lancet Neurol 15, 455–532.

Ayton

, Lei

, Bush

(2013) Metallostasis in Alzheimer’s disease. Free Radic Biol Med 62, 76–89.

Tang

, Oliveros

, Jang

(2019) Dysfunctional mitochondrial bioenergetics and synaptic degeneration in Alzheimer disease. Int Neurourol J 23, S5–10.

Ahmad

(2013) Overlapped metabolic and therapeutic links between Alzheimer and diabetes. Mol Neurobiol 47, 399–424.

Ozben

, Ozben

(2019) Neuro-inflammation and anti-inflammatory treatment options for Alzheimer’s disease. Clin Biochem 72, 87–89.

Adlard

, Bush

(2012) Metal chaperones: A holistic approach to the treatment of Alzheimer’s disease. Front Psychiatry 3, 15.

Arbiser

, Kraeft

, van Leeuwen

, Hurwitz

, Selig

, Dickersin

, Flint

, Byers

, Chen

(1998) Clioquinol-zinc chelate: a candidate causative agent of subacute myelo-optic neuropathy. Mol Med 4, 665–670.

Villemagne

, Rowe

, Barnham

, Cherny

, Woodward

, Bozinosvski

, Salvado

, Bourgeat

, Perez

, Fowler

, Rembach

, Maruff

, Ritchie

, Tanzi

, Masters

(2017) A randomized, exploratory molecular imaging study targeting amyloid beta with a novel 8-OH quinoline in Alzheimer’s disease: The PBT2-204 IMAGINE study. Alzheimers Dement (N Y) 3, 622–635.

10.

Yassin

, Ekblom

, Xilinas

, Gottfries

, Oreland

(2000) Changes in uptake of vitamin B(12) and trace metals in brains of mice treated with clioquinol. J Neurol Sci 173, 40–44.

11.

Pavlov

, Hutter-Paier

, Havas

, Windisch

, Winblad

(2018) Development of GMP-1 a molecular chaperone network modulator protecting mitochondrial function and its assessment in fly and mice models of Alzheimer’s disease. J Cell Mol Med 22, 3464–3474.

12.

Devi

, Prabhu

, Galati

, Avadhani

, Anandatheerthavarada

(2006) Accumulation of amyloid precursor protein in the mitochondrial import channels of human Alzheimer’s disease brain is associated with mitochondrial dysfunction. J Neurosci 26, 9057–9068.

13.

Barnham

, Bush

(2014) Biological metals and metal-targeting compounds in major neurodegenerative diseases. Chem Soc Rev 43, 6727–6749.

14.

Di Vaira

, Bazzicalupi

, Orioli

, Messori

, Bruni

, Zatta

(2004) Clioquinol, a drug for Alzheimer’s disease specifically interfering with brain metal metabolism: structural characterization of its zinc(II) and copper(II) complexes. Inorg Chem 43, 3795–3797.

15.

Ferrada

, Arancibia

, Loeb

, Norambuena

, Olea-Azar

, Huidobro-Toro

(2007) Stoichiometry and conditional stability constants of Cu(II) or Zn(II) clioquinol complexes; implications for Alzheimer’s and Huntington’s disease therapy. Neurotoxicology 28, 445–449.

16.

Borghesani

, Alies

, Hureau

(2018) Cu(II) binding to various forms of amyloid-beta peptides. Are they friends or foes? Eur J Inorg Chem 2018, 7–15.

17.

Rozga

, Kloniecki

, Dadlez

, Bal

(2010) A direct determination of the dissociation constant for the Cu(II) complex of amyloid beta 1-40 peptide. Chem Res Toxicol 23, 336–340.

18.

Bush

, Pettingell

Jr. , Paradis

, Tanzi

(1994) Modulation of A beta adhesiveness and secretase site cleavage by zinc. J Biol Chem 269, 12152–12158.

19.

Matlack

, Tardiff

, Narayan

, Hamamichi

, Caldwell

, Lindquist

(2014) Clioquinol promotes the degradation of metal-dependent amyloid-beta (Abeta) oligomers to restore endocytosis and ameliorate Abeta toxicity. Proc Natl Acad Sci U S A 111, 4013–4018.

20.

Mold

, Ouro-Gnao

, Wieckowski

, Exley

(2013) Copper prevents amyloid-beta(1-42) from forming amyloid fibrils under near-physiological conditions in vitro. Sci Rep 3, 1256.

21.

Garzon-Rodriguez

, Yatsimirsky

, Glabe

(1999) Binding of Zn(II), Cu(II), and Fe(II) ions to Alzheimer’s A beta peptide studied by fluorescence. Bioorg Med Chem Lett 9, 2243–2248.

22.

El Khoury

, Hickman

, Thomas

, Cao

, Silverstein

, Loike

(1996) Scavenger receptor-mediated adhesion of microglia to beta-amyloid fibrils. Nature 382, 716–719.

23.

Behl

, Davis

, Lesley

, Schubert

(1994) Hydrogen peroxide mediates amyloid beta protein toxicity. Cell 77, 817–827.

24.

Yang

, Wang

, Chen

, Wang

, Zheng

(2018) A comparative study of resveratrol and resveratrol-functional selenium nanoparticles: Inhibiting amyloid beta aggregation and reactive oxygen species formation properties. J Biomed Mater Res A 106, 3034–3041.

25.

Sharma

, Pavlova

, Kim

, Mirica

(2013) The effect of Cu(2+) and Zn(2+) on the Abeta42 peptide aggregation and cellular toxicity. Metallomics 5, 1529–1536.

26.

Bush

, Tanzi

(2008) Therapeutics for Alzheimer’s disease based on the metal hypothesis. Neurotherapeutics 5, 421–432.

27.

Santos

, Chand

, Chaves

(2016) Recent progress in multifunctional metal chelators as potential drugs for Alzheimer’s disease. Coord Chem Rev 327-328, 287–303.

28.

Lang

, Fan

, Wang

, Zheng

, Xiao

, Wang

, Zhong

, Zhou

(2013) Inhibition of human high-affinity copper importer Ctr1 orthologous in the nervous system of Drosophila ameliorates Abeta42-induced Alzheimer’s disease-like symptoms. Neurobiol Aging 34, 2604–2612.

29.

Lang

, Wang

, Fan

, Xiao

, Wang

, Zhong

, Zhou

(2012) Genetic inhibition of solute-linked carrier 39 family transporter 1 ameliorates abeta pathology in a Drosophila model of Alzheimer’s disease. PLoS Genet 8, e1002683.

30.

Hua

, Munter

, Harmeier

, Georgiev

, Multhaup

, Schaffner

(2011) Toxicity of Alzheimer’s disease-associated Abeta peptide is ameliorated in a Drosophila model by tight control of zinc and copper availability. Biol Chem 392, 919–926.

31.

Kumar

, Tiwari

, Sharma

(2018) Changing paradigm from one target one ligand towards multi-target directed ligand design for key drug targets of Alzheimer disease: an important role of in silico methods in multi-target directed ligands design. Curr Neuropharmacol 16, 726–739.

32.

Shamloo

, Asadbegi

, Khandan

, Amanzadi

(2018) Designing a new multifunctional peptide for metal chelation and Abeta inhibition. Arch Biochem Biophys 653, 1–9.

33.

Sharma

, Pavlova

, Kim

, Finkelstein

, Hawco

, Rath

, Kim

, Mirica

(2012) Bifunctional compounds for controlling metal-mediated aggregation of the abeta42 peptide. J Am Chem Soc 134, 6625–6636.

34.

Braymer

, Choi

, DeToma

, Wang

, Nam

, Kampf

, Ramamoorthy

, Lim

(2011) Development of bifunctional stilbene derivatives for targeting and modulating metal-amyloid-beta species. Inorg Chem 50, 10724–10734.

35.

Choi

, Braymer

, Nanga

, Ramamoorthy

, Lim

(2010) Design of small molecules that target metal-Abeta species and regulate metal-induced Abeta aggregation and neurotoxicity. Proc Natl Acad Sci U S A 107, 21990–21995.

36.

Hilt

, Altman

, Kalai

, Maezawa

, Gong

, Wachsmann-Hogiu

, Jin

, Voss

(2018) A bifunctional anti-amyloid blocks oxidative stress and the accumulation of intraneuronal amyloid-beta. Molecules 23, E2010.

37.

Marinelli

, Fornasari

, Di Stefano

, Turkez

, Arslan

, Eusepi

, Ciulla

, Cacciatore

(2017) (R)-alpha-Lipoyl-Gly-l-Pro-l-Glu dimethyl ester as dual acting agent for the treatment of Alzheimer’s disease. Neuropeptides 66, 52–58.

38.

Reis

, Cagide

, Valencia

, Teixeira

, Bagetta

, Perez

, Uriarte

, Oliveira

, Ortuso

, Alcaro

, Rodriguez-Franco

, Borges

(2018) Multi-target-directed ligands for Alzheimer’s disease: Discovery of chromone-based monoamine oxidase/cholinesterase inhibitors. Eur J Med Chem 158, 781–800.

39.

Jiang

, Chen

, Zhou

, He

, Yang

, Chen

, Qu

, Feng

, Sun

(2018) Dual GSK-3beta/AChE inhibitors as a new strategy for multitargeting anti-Alzheimer’s disease drug discovery. ACS Med Chem Lett 9, 171–176.

40.

Rampa

, Gobbi

, Concetta Di Martino

, Belluti

, Bisi

(2017) Dual BACE-1/GSK-3beta inhibitors to combat Alzheimer’s disease: a focused review. Curr Top Med Chem 17, 3361–3369.

41.

Di Martino

, De Simone

, Andrisano

, Bisignano

, Bisi

, Gobbi

, Rampa

, Fato

, Bergamini

, Perez

, Martinez

, Bottegoni

, Cavalli

, Belluti

(2016) Versatility of the curcumin scaffold: discovery of potent and balanced dual BACE-1 and GSK-3beta inhibitors. J Med Chem 59, 531–544.

Metal Binding by GMP-1 and Its Pyrimido [1,2]benzimidazole Analogs Confirms Protection Against Amyloid-β Associated Neurotoxicity

Abstract

Keywords

INTRODUCTION

METHODS

Chemicals

Determination of the metal to GMP-1 binding and stoichiometry by UV/Vis spectroscopy

Limited proteolysis of β-casein

Cytochrome oxidase activity assay

Cellular viability assays

Reactive oxygen species (ROS) detection using DCFDA

Aβ1–42 fibrillization assay

RESULTS

Metal binding properties of GMP-1

Effect of GMP-1 versus metal chelators on metal-dependent enzymatic activities

Effect of GMP-1 on Aβ1-42 copper binding and copper-dependent Aβ1-42 aggregation

Prevention of Aβ1-42–Cu(II) complexes-induced ROS generation by GMP-1 and CQ

GMP-1 and CQ inhibit SH-SY5Y neuroblastoma cell death treated with Aβ1–42–Cu(II) complexes

DISCUSSION

Footnotes

ACKNOWLEDGMENTS

References

Aβ_1–42 fibrillization assay

Effect of GMP-1 on Aβ₁_-₄₂ copper binding and copper-dependent Aβ₁_-₄₂ aggregation

Prevention of Aβ_1-42–Cu(II) complexes-induced ROS generation by GMP-1 and CQ

GMP-1 and CQ inhibit SH-SY5Y neuroblastoma cell death treated with Aβ_1–42–Cu(II) complexes