8-Hydroxyquinolylnitrones as multifunctional ligands for the therapy of neurodegenerative diseases

Abstract

We describe the development of quinolylnitrones (QNs) as multifunctional ligands inhibiting cholinesterases (ChEs: acetylcholinesterase and butyrylcholinesterase–hBChE) and monoamine oxidases (hMAO-A/B) for the therapy of neurodegenerative diseases. We identified QN 19, a simple, low molecular weight nitrone, that is readily synthesized from commercially available 8-hydroxyquinoline-2-carbaldehyde. Quinolylnitrone 19 has no typical pharmacophoric element to suggest ChE or MAO inhibition, yet unexpectedly showed potent inhibition of hBChE (IC₅₀ = 1.06 ± 0.31 nmol/L) and hMAO-B (IC₅₀ = 4.46 ± 0.18 μmol/L). The crystal structures of 19 with hBChE and hMAO-B provided the structural basis for potent binding, which was further studied by enzyme kinetics. Compound 19 acted as a free radical scavenger and biometal chelator, crossed the blood–brain barrier, was not cytotoxic, and showed neuroprotective properties in a 6-hydroxydopamine cell model of Parkinson's disease. In addition, in vivo studies showed the anti-amnesic effect of 19 in the scopolamine-induced mouse model of AD without adverse effects on motoric function and coordination. Importantly, chronic treatment of double transgenic APPswe-PS1δE9 mice with 19 reduced amyloid plaque load in the hippocampus and cortex of female mice, underscoring the disease-modifying effect of QN 19.

Graphical abstract

Quinolylnitrone 19 was identified as a dual hBChE/hMAO-B inhibitor that improved cognition in scopolamine-induced mice model and reduced amyloid plaque load in cortex and hippocampus of APPswe/PS1δE9 mice.

1. Introduction

Multifactorial and complex diseases e.g., neurodegenerative and malignant diseases, can be efficiently treated by multifunctional ligands targeting multiple biological targets involved in disease onset and progression^1,2. The design of such ligands is usually based on the combination of structurally overlapping elements, or pharmacophoric moieties separated by linkers, which are directed against individual targets^3,4. The fact that multifunctional ligands are not yet suited for the clinical therapy of Alzheimer's disease (AD), demonstrates that their development is not a trivial task. Nevertheless, this approach remains a valuable therapeutic strategy to pursue due to the numerous advantages pointed out in the literature^1,5,6.

AD is a deleterious neurodegenerative disease with an archetypal gradual decline in memory and learning capacity as a result of a number of pathological alterations in the central nervous system (CNS)⁷. Even frontline tau and amyloid β hypotheses of AD have yet to provide an efficient and safe anti-AD therapy that could be widely accepted and used in the clinic, despite the recent FDA's approval of aducanumab⁸. Aducanumab reduced brain Aβ load and slowed cognitive decline⁹, but failed to replicate this effect in a larger cohort¹⁰. The failure may have been due to the disease heterogenicity, underscoring the need for careful stratification of patients and clear inclusion criteria¹¹. The literature review shows that inhibition of monoamine oxidases (MAO-A/B), β-secretase and cholinesterases (ChEs, namely acetylcholinesterase (AChE) and butyrylcholinesterase (BChE)) in addition to modulation of monoaminergic receptors remains the focus of AD-related small molecule drug design in academia12, 13, 14. A reduction in acetylcholine (ACh) levels is caused by the death of cholinergic neurons in certain brain areas that mediate cognition¹⁵. Therefore, the inhibition of AChE or BChE increases ACh levels at the synaptic cleft and restores cholinergic neurotransmission¹⁶. MAOs, on the other hand, oxidize neurotransmitter-functioning amines, hence generating H₂O₂ and reactive nitrogen and oxygen species¹⁷. MAO inhibition confers substantial neuroprotective benefits by reducing oxidative stress, and improves memory, learning and damaged synaptic plasticity in mouse model of AD through regulating tonic GABA levels18, 19, 20, 21. Parkinson's disease (PD), which affects 1% of adults over 60, is another chronic neurodegenerative condition defined by gradual death of dopaminergic neurons and the development of Lewy bodies with α-synuclein aggregates²². Bradykinesia, tremor and postural instability, among others, are all rotted in a deficiency of dopaminergic neurotransmission. Approved medicines for PD boost dopamine in striatum by selectively blocking MAO-B. Rasagiline, for instance, is used alone in the early stages of PD and as add-on therapy to levodopa in the later stages of the disease²³. A recently discovered reversible MAO-B inhibitor was able to repair memory impairment and learning in APP/PS1 mouse model of AD. This gives a solid argument to examine MAO-B inhibitors as potential treatments for AD²⁴.

Quinoline is a privileged heterocycle and the central core of series of compounds that exhibit various diverse biological activities²⁵. Of particular interest are 8-hydroxyquinolines, which have been the subject of continued medicinal chemistry research over the past decades, resulting in the discovery of new ligands^26,27 such as HLA20 (Fig. 1)²⁸, a brain permeable chelator that shows selective inhibition of human (h)MAO-B and significant neuroprotective activity against 6-hydroxydopamine (6-OHDA)-induced death in differentiated P19 cells. Another example is donepezil/propargylamine/8-hydroxyquinoline hybrid DPH6 (Fig. 1)²⁹, an irreversible hMAO-A/B inhibitor as well as mixed-type hAChE inhibitor that has metal-chelating capabilities³⁰. Building on the 8-hydroxyquinoline core, neuroprotective quinolylnitrone (QN) QN23 was developed (Fig. 1)³¹, which was active in vivo animal models of global and localized cerebral ischemia. Contilisant (Fig. 1)³² is a brain-permeable neuroprotective agent that inhibits hChEs and hMAOs, modulates histamine H₃ and σ₁ receptors, and outperformed donepezil in animal models of AD³³. Juxtaposition of QN23 and Contilisant yielded QN hybrid MC903 (Fig. 1), which showed potent neuroprotective properties in cell-based models of ischemia and AD³⁴.

Figure 1.

Structures of quinolines HLA20, DPH6, QN23, Contilisant and MC903.

With this in mind, and based on the structure and neuroprotective activity of QN MC903 (Fig. 1)³⁴, we designed, synthesized, and evaluated QNs with general formulas I and II by attaching pharmacophoric moieties of Contilisant and MC903, namely N-propylpiperidine and 1-propargyl-4-propylpiperazine, onto a 8-hydroxyquinoline core (Fig. 2). These efforts allowed us to identify QN 19 as a readily available nitrone, that potently inhibited hBChE and hMAO-B, was a potent antioxidant and metal chelator protecting cells in 6-hydroxydopamine-induced cell model of PD, improved cognition in scopolamine-induced amnesic mice and importantly, reduced Aβ burden in cortex and hippocampus of female double transgenic APPswe-PS1δE9 mice.

Figure 2.

General structures of QNs I and II.

2. Results and discussion

2.1. Chemistry

QNs I (1–12, Table 1) and QNs II (13–19, Table 2) (Fig. 2) with tert-butyl- or benzyl-nitrone at C3 or C2 positions on the 8-hydroxyquinoline core, respectively, were prepared by reacting the corresponding quinolylcarbaldehydes with commercially available N-alkylhydroxylamines (Scheme 1, Scheme 2). For comparison, oxime 33 (Table 2)³⁵ was synthesized from 8-hydroxyquinoline-2-carbaldehyde 30.

Table 1.

MAO and ChE inhibition of QNs 1–12.

QN	Structure	IC50 ± SEM (μmol/L)a
		hMAO-A	hMAO-B	hAChE	hBChE
MC903		/b	/b	/b	8.38 ± 0.75
1		/b	/b	/b	0.7074 ± 0.0454
2		/b	/b	39.6 ± 6.1	0.4153 ± 0.0167
3		/b	/b	/b	2.06 ± 0.04
4		/b	/b	/b	3.06 ± 0.29
5		112.8 ± 13.2	/b	44.7 ± 4.8	5.57 ± 0.46
6		105.7 ± 3.7	50.4 ± 4.5	68.6 ± 13.6	10.2 ± 1.1
7		/b	/b	/b	13.3 ± 0.9
8		14.9 ± 1.2	13.8 ± 1.8	/b	14.8 ± 2.1
9		/b	/b	/b	15.2 ± 1.1
10		/b	/b	/b	/b
11		/b,c	/b,c	/b	12.4 ± 0.7
12		/b	/b	/b	10.7 ± 0.7

^aIC₅₀ reported as average ± SEM (n = 3, in triplicates).

^bnot active (RA at 100 μmol/L ≥ 50%).

^cNonspecific inhibition at 100 μmol/L due to solubility issues, no inhibition at 10 μmol/L.

Table 2.

MAO and ChE inhibition of QNs 13–19, oxime 33, aldehydes 30 and 32, and positive controls.

QN	Structure	IC50 ± SEM (μmol/L)a
		hMAO-A	hMAO-B	hAChE	hBChE
13	4	/b	/b	/b	0.0254 ± 0.0021
14		/b	/b	/b,c	0.0769 ± 0.0063
15		/b	/b	/b	0.0073 ± 0.0011
16		/b,c	/b	/b	0.0310 ± 0.0064
17		/b,c	10.1 ± 2.5	/b	/b
18		/b	16.4 ± 1.0	/b	0.0051 ± 0.0006
19		/b	4.46 ± 0.18	0.0290 ± 0.0030	0.0011 ± 0.00031d
30		/b	43.9 ± 3.4	/b	5.26 ± 0.17
32		/b	/b	/b	0.0015 ± 0.0003d
33		/b	/b	/b	/b
PBN		/b	/b	/b	/b
Donepezil		n.d.e	n.d.e	0.0220 ± 0.0024	4.15 ± 0.56
Safinamide		/b	0.029 ± 0.002	n.d.e	n.d.e

^aIC₅₀ reported as average ± SEM (n = 3, in triplicates).

^bn. a., not active (RA at 100 μmol/L ≥ 50%).

^cNonspecific inhibition at 100 μmol/L due to solubility issues, no inhibition at 10 μmol/L.

^dIC₅₀ value approaching the hBChE concentration (approx. 1 nmol/L) in the in vitro assays.

^en.d., not determined.

Scheme 1.

Syntheses of QNs 1–12. Conditions: (a) corresponding primary alcohol, t-BuOK, NaI, 1,4-dioxane, 100 °C, 6 h, MWI (21, 74%; 23, 69%); (b) HCl_(aq), THF, rt, 1 h (22, 24, 99%); (c) corresponding N-alkylhydroxylamine hydrochloride, Na₂SO₄, NaHCO₃, THF, 90 °C, 1.5–4.5 h, MWI (1–4, 58%–94%); (d) BBr₃, DCM, −78 °C→ rt, 3 h (75%); (e) corresponding alkyl chloride, K₂CO₃, CHCl₃, H₂O, (cat. KI for 29), 80–85 °C, 24–48 h (27, 65%; 29, 83%); (f) corresponding N-alkylhydroxylamine hydrochloride, Na₂SO₄, NaOAc, EtOH, 90 °C, 2–3 h, MWI (5–12, 35%–99%).

Scheme 2.

Syntheses of QNs 13–19 and oxime 30. Conditions: (a) corresponding alkyl chloride, K₂CO₃, CHCl₃, H₂O, 80 °C, 24–48 h (31, 89%; 32, 53%); (b) corresponding N-alkylhydroxylamine hydrochloride, Na₂SO₄, NaOAc, EtOH, 90–95 °C, 10 min–3 h, MWI (13–19, 52%–98%).

All new compounds were characterized by their analytical and spectroscopic data (for details see Experimental Section and Supporting Information), which agreed well with their structures. In particular, the QNs were obtained as Z-isomers³⁶, showing ¹H and ¹³C NMR data consistent with those expected and previously demonstrated for related quinolinenitrones³¹. Only in the case of N-benzylquinolylnitrones 14 and 16, we detected a Z and E isomers mixture in 3.5 to 1, and one to 1.5 ratios, respectively, that could not be separated by column chromatography, and were biologically evaluated as such. Stereochemical assignment was tentatively proposed on the basis that the majority of pure (Z)-N-benzylnitrones described showed singlets for the methylene group of the benzyl motif in the range of chemical shifts in 5.23–5.16 ppm range. For example, major (Z)-14 isomer showed the N(O)–CH₂C₆H₅ signal at δ = 5.15 ppm, while the major (E)-16 isomer showed this signal at δ = 5.44 ppm.

2.2. Pharmacological evaluation

The multifunctional properties of the synthesized QNs were investigated in bottom–up approach, in which analogues were first screened for their MAO and ChEs inhibitory potencies. A selection of active compounds was progressed into next level of evaluation that encompassed evaluation of antioxidative properties in the DPPH assay, the capacity to inhibit amyloid β_1–42 (Aβ_1–42) aggregation, prediction of blood–brain barrier (BBB) permeation, and chelation of metal ions. Lastly, cell-based assays and in vivo studies were undertaken for the most promising compound.

2.2.1. Inhibition of ChEs and MAOs

First round evaluation of QNs encompassed the study of their inhibitory potencies against MAOs using horseradish-Amplex Red coupled assay³⁷ and both ChEs, namely hAChE and hBChE using Ellman's method (Table 1, Table 2)³⁸.

Parent MC903 showed only single-digit micromolar hBChE inhibition that was generally retained regardless of the major structural changes of type I QNs that all share either benzyl- or tert-butyl-nitrone moiety at position three of quinoline (Table 1). A noticeable increase in hBChE inhibitory potency was seen for compounds 1 and 2, both being 1-propyloxypiperidine derivatives. Only modest micromolar inhibition of hAChE and hMAO-A/B was determined for certain compounds, i.e., 2, 5, 6, and 8.

An intriguing improvement of inhibition was seen for type II QNs 13–19, where the nitrone pharmacophore was attached at position two of 8-hydroxyquinoline. Pairwise comparisons of type I (1–4) and type II QNs (13–16) revealed that the latter were up to 1000-fold more potent hBChE inhibitors. Remarkable was also the selectivity of QNs 13–16, as none of them inhibited the structurally related hAChE. This may have occurred because hAChE's smaller active site creates steric hindrance compared to that of hBChE³⁹. Truncation of the 8-alkyloxy substituent generated compounds 18 and 19, and to our surprise further lowered IC₅₀ values. On the other hand, replacement of benzylnitrone or tert-butylnitrone with methylnitrone in 17 or oxime in 33 completely abolished hBChE inhibition. Further structure reduction to 8-hydroxyquinoline-2-carbaldehyde 30 regained hBChE inhibition with IC₅₀ of 5.26 ± 0.17 μmol/L. The latter prompted us to also assay aldehyde 32, which potently and selectively inhibited hBChE. None of the QNs 13–19, neither oxime 33, nor aldehydes 30 and 32 inhibited hMAO-A, and only smaller 8-hydroxyquinolines 17–19 inhibited hMAO-B in low micromolar concentrations. One hundred-fold dilution assay showed that hMAO-B inhibition by 19 was reversible (Supporting Information Fig. S1).

To the best of our knowledge, this is only the second instance that nitrones are reported as hMAO-B and hBChE inhibitors. MT-20 R (Fig. 3) is a hybrid molecule designed by merging the propargylamine–MAO pharmacophore, and N-methyl-N-ethyl carbamate–ChE pharmacophore, on α-phenyl-tert-butylnitrone (PBN) core⁴¹. MT-20 R is an efficient radical scavenger that inhibits both MAO-B and AChE, and is active in vivo PD models⁴⁰. The synthesis of (R)-MT-20 R and analogues has not been reported, only the preparation of the racemic form and MAO-B inhibitory potencies (MAO-B: IC₅₀ = 1.05 ± 0.16 μmol/L) were disclosed in a patent⁴¹. In comparison, nitrone 19 has several advantages: a) it is structurally simple, and readily synthetically available; b) it is devoid of chiral centers; and c) it is a reversible hMAO-B inhibitor. The latter may be of importance because the potential off-target effects, the hepatotoxic–mutagenic profile described for irreversible propargylamine derivatives, and the potential immunogenicity of the resulting target adduct leading to allergic reactions are the major concerns that have prevented the development of MAO-B inhibitors of the propargylamine class⁴².

Figure 3.

Structures of compounds 19, MT-20 R and 34.

The inhibitory potencies of QNs on ChEs have also been poorly studied and documented. For example, (i) PBN inhibits murine AChE (K_i = 0.58 mmol/L; K_i′ = 2.99 mmol/L)⁴³; (ii) the cardiovascular drug chlordiazepoxide inhibits Electrophorus electricus AChE (eeAChE) and horse serum BChE (eqBChE) with K_i values of 50 ± 7 μmol/L and 0.30 ± 0.04 μmol/L, respectively⁴⁴; (iii) a series of nitrone-lipoic acid derivatives inhibit AChE with IC₅₀ = 19.1–278.8 μmol/L, and BChE with IC₅₀ = 46.4–247.2 μmol/L⁴⁵; (iv) 2-benzazepine nitrones, which are able to protect dopaminergic neurons from 6-OHDA-induced oxidative toxicity, barely inhibit eeAChE, but show modest inhibition of eqBChE with IC₅₀ = 16.8–418 μmol/L, compared with PBN (eqBChE: IC₅₀ = 478 ± 38 μmol/L), which showed no inhibition of eeAChE⁴⁶ and hAChE (Table 2); (v) benzoic acid-derived nitrones selectively inhibit eeAChE, with nitrone 34 (Fig. 3) being the most potent (IC₅₀ = 8.3 ± 0.3 μmol/L)⁴⁷.

2.2.2. Crystal structure of hBChE with QNs 15, 19 and 32

Crystal structures of hBChE in complex with 15, 19 and 32 were determined after crystal soaking in crystallization buffer containing 2.5 mmol/L of the respective ligand (Supporting Information Table S1). Beside extra electron density close to specific asparagine residues corresponding to N-glycosylation sites, and at the protein surface for ions and other ligands, i.e., glycerol and buffer, positive and continuous electron densities were observed in the active site (F_o–F_c map, Supporting Information Fig. S2). For 15, one molecule was fitted in the electron density (Fig. 4A). The substituted 8-hydroxyquinoline ring was oriented towards Phe329 and Trp231, forming a T-shaped π–π interaction, respectively, 4.8 and 4.1 Å from Trp231 and Phe329. The tert-butyl group oriented towards Trp82, while the piperazine ring folded back over the 8-hydroxyquinoline ring and interacted with Tyr332 at a ring to nitrogen distance of 4.1 Å. The nitrone oxygen was at hydrogen bond distance from a conserved water molecule (3.1 Å). For the shorter QN 19, two molecules of ligand were fitted in the electron density (Fig. 4B). The molecule deepest in the active site gorge had the 8-hydroxyquinoline ring engaged in a T-shaped π–π interaction with both Trp231 and Phe329, with ring distances of 3.6 Å and 3.8 Å respectively, while the hydroxyl group pointed towards the catalytic residues (2.7 Å from Ser198-Oγ and 3.5 Å from His438-Nε). On the other hand, the benzyl moiety formed a planar π–π interaction with Trp82 (mean ring distance of 3.7 Å), and an additional T-shaped π–π interaction with the 8-hydroxyquinoline ring of the second 19 molecule at a distance of 3.3 Å. The nitrone oxygen did not engage in a specific interaction with hBChE being at 4.0 Å from Thr120-Oγ; however, it more likely formed a π–π interaction with the 8-hydroxyquinoline of the second 19 molecule bound (interplanar distance of 3.6 Å). Additionally, the 8-hydroxyquinoline ring of the second molecule was fitted in an aromatic cavity formed by Trp82, Phe329, Tyr332, Trp430, and Tyr440, while the benzyl ring did not form any specific interaction (closest distance of 3.4 Å to Thr120-Cγ). This was supported by the higher mean B-factor of the benzyl ring atoms (79.2 Å²) compared to those of the 8-hydroxyquinoline ring (67.9 Å²). Binding of two inhibitor molecules into hBChE active site was already described for thioflavin T (ThT)⁴⁸. Finally, in the case of 32, the electron density defining the ligand presented continuity with the one of the residues–Ser198 (Fig. 4C). Ligand fitting resulted in a position of the carbonyl carbon of 32 about 1.4 Å distance of Ser198-Oγ, in accordance with covalent bonding. This can be explained by the reaction of Ser198-Oγ with the aldehyde of 32 to form a hemiacetal. Besides reaction with organophosphorous or carbamates compounds, such covalent modification of the catalytic serine has been previously reported in Torpedo AChE in complex with ketone-containing ligands such as (N,N,N-trimethylammonio)-2,2,2-trifluoroacetophenone⁴⁹ or 4-oxo-N,N,N-trimethylpentanaminium⁵⁰ used as non-hydrolysable substrate analogs. Specific bond distance, angles and dihedral values were defined during refinement using values calculated in Phenix eLBOW for a serine residue hemiacetylated with compound 32. The 8-hydroxyquinoline ring reoriented in a position similar to 15 with a small deviation (24–25°) and engaged a T-shaped π–π interactions with both Trp231 and Phe329 with respective distances of 3.6 and 4.3 Å. The hydroxyl of the hemiacetal was stabilized in the oxyanion hole (Gly116, Gly117 and Ala199), with respective distances of 2.7, 2.8 and 2.7 Å. Compared to 15, the piperazine ring slid away from Tyr332. In this position, 32 specifically trapped a water molecule with specific H-bonds between the oxygen and nitrogens atoms of the 8-hydroxyquinoline (3.1 and 3.3 Å, respectively), and the proximal nitrogen of the piperazine with a distance of 2.9 Å. The distal nitrogen of the piperazine is at H-bond distance from a water molecule hydrogen-bonded to Tyr332-OH (2.6 and 2.4 Å, respectively). Some additional electron density corresponding to a small ligand was present in the vicinity of the piperidine ring and Trp82. Such electron density is often observed in hBChE structures and was modeled here as a propionate hydrogen-bonded to Glu 197 (2.6 Å). Overall, interaction complementarity of ligands in the hBChE active site provide the molecular basis for potent inhibition of hBChE.

Figure 4.

X-ray structures of (A) 15 and (B) 19 and (C) 32 bound to hBChE. hBChE residues in interaction with inhibitors are represented as green sticks, and active site as a light grey surface. Calculated polder maps⁵¹ around ligands bound are represented as dark grey mesh. 15, 19 and 32 are shown as sticks (carbons in cyan, orange/magenta and yellow, respectively, nitrogens in blue and oxygens in red), H-bonds as dashed lines and water molecules as red spheres.

2.2.3. Crystal structure and molecular dynamics of hMAO-B in complex with 19

The structure of hMAO-B with 19 was solved at 2.0 Å (PDB entry 7ZW3) after co-crystallization with 1 mmol/L 19 (Supporting Information Table S2). The electron density of 19 was slightly better in chain B with respect to A, even though the conformation of the inhibitor bound to the active site was the same (Supporting Information Fig. S3). Therefore, we will subsequently refer to chain B for the description of the structure. The electron density was continuous and allowed the unambiguous fitting of the ligand (Fig. 5).

Figure 5.

X-ray structure of 19/hMAO-B. hMAO-B (chain B) active site residues are represented as green sticks, and cavity as a light grey semi-transparent surface. The refined 2F_o–F_c (dark grey mesh) was improved by applying FEM algorithm (Phenix)⁵³. 19 is shown as sticks (carbon atoms in orange, oxygens and nitrogens in red and blue, respectively) and H-bonds as dashed lines.

The 8-hydroxyquinoline ring was bound in front of the isoalloxazine ring of the flavin with the hydroxyl group pointing upward near Phe343. The benzyl moiety was rotated of about 90° with respect to the 8-hydroxyquinoline ring and lain in the hydrophobic entrance cavity. The hMAO-B hydrophobic cavity imposes 19 to adopt an extended structure (with Ile199 in open state), whereas in hBChE it binds in a more folded conformation. The weaker electron density at the level of the nitrone moiety suggested a certain degree of flexibility of the inhibitor within the entrance cavity of hMAO-B. This was supported by the higher mean B-factor values of the benzylnitrone chain (66.2 Å²) with respect to the 8-hydroxyquinoline ring (45.3 Å²). Moreover, Cys172 was found in a double conformation, with occupancy 0.7 for the rotamer that is hydrogen-bonded to the ligand (Cys172 sulfur-inhibitor oxygen distance = 4.2 Å), suggesting that the position of the nitrone in the final refined model corresponds to the preferential binding mode of the ligand. The Cys172 double conformation was previously found in structures in complex with inhibitors that display a hydrogen bond acceptor moiety in this position⁵².

The analysis of the 500 ns MD simulation revealed that 19 mainly interacted with the hMAO-B active site through hydrophobic interactions, including short-lived π–π interactions, and transient water bridges with the 8-hydroxy and nitrone moiety. After 105 ns, the 8-hydroxyquinoline moiety underwent a vertical flip, and was involved in transient interactions with three tyrosines (Tyr188, 398, 435) in the vicinity of isoalloxazine moiety. The nitrone's oxygen was predominantly hydrogen-bonded to Ile198's carbonyl through a water bridge (Fig. 6 and Supporting Information Fig. S4), and not to Cys172, as observed in the crystal structure (9.8 Å mean distance between S^Cys172–O^nitrone during MD). The higher mean B-factor values, indicating the higher flexibility, of the N-benzyl sidechain in comparison to the 8-hydroxyquinoline were also reflected in almost twice larger per-atom RMSF values during MD simulation (Supporting Information Figs. S5 and S6). Upon visual inspection of the trajectory, the benzyl sidechain often flipped its position, which corresponds with the weaker electron density in the crystal structure.

Figure 6.

Diagram showing the evolution of protein-ligand contacts diagram 19-hMAO-B complex during molecular dynamic simulation. Interactions occurring >10% of the simulation time (i.e., the frames) are shown. The hydrogen bonds (also in water bridges) are shown as violet lines and hydrophobic interactions in green circles, while grey circles denote solvent exposure.

2.2.4. Kinetic evaluation

Further in vitro kinetic evaluation was performed on compound 19 for hBChE and for hMAO-B. Inhibition of hBChE by 19 was studied by obtaining progress curves of the substrate hydrolysis (butyrylthiocholine iodide, BTCI) by recombinant hBChE in the absence (red lines) and presence of 19 (yellow lines) (Fig. 7A). For experiments in Fig. 7A using varying concentrations of 19, the yellow curves were started by the addition of enzyme, while the blue curves were started with the incubated mixture of hBChE and 19 that had been previously preincubated for 20 min. The analysis of the curves showed that there was a good agreement between the theoretical model (black lines) and experimental data (yellow, red and blue lines) that defined reaction mechanism and binding of two 19 molecules to both free, substrate bound and acylated (i.e., butylated, ES) hBChE, as presented in model (Fig. 7B)⁵⁴. The fit converged with nearly perfect accordance with the experimental curves, thus the standard errors of all constants were less the one percent (Fig. 7C). The characteristic kinetic parameters (i.e., K_m, k_cat) for the interaction of BTCI with hBChE were determined previously and were fixed in the analysis⁵⁵. If the substrate, i.e., BTCI, was already bound at the bottom of the active site, then 19 could only bind to the peripheral anionic site (IES), where the low nanomolar dissociation constant K_i of 4.3 nmol/L reflected the binding to free and substrate bound hBChE⁵⁶. Initial binding of the inhibitor to the peripheral site was followed by the slow isomerization steps (i.e., k_iso+, k_iso–)–the sliding and accommodation of 19 to the bottom of the active site. The dissociation constant for the binding of the second molecule of QN 19 to hBChE is denoted by symbol K_ii. The binding of the first inhibitor molecule to the active site then enabled the second molecule to accommodate into the active site with K_ii of 14.5 nmol/L. The kinetic data in solution with the low nanomolar concentration of inhibitor 19 confirmed the structural data with two inhibitor molecules bound.

Figure 7.

Inactivation of hBChE and hMAO-B by 19. (A) Parabolic inhibition and reactivation of hBChE by 19 at 50 μmol/L BTCI, 1 mmol/L DTNB and 0.5 nmol/L hBChE. i) 19 concentrations were zero (red curve), 1, 2, 3, 4, 5 and 7.5 nmol/L (yellow curves). ii–iv) 19 at 0 nmol/L (red curve) and ii) 1 nmol/L, iii) 2 nmol/L, and iv) 3 nmol/L concentration without (yellow) and with 20 min preincubation with hBChE. (B) The scheme for the interaction between hBChE and 19. E, free enzyme; I, 19; IE, instantaneously bound 19 to the peripheral site; EI, 19 bound to the bottom of catalytic site; IEI, fully occupied active site as observed by crystallography (Fig. 4B). (C) Kinetics for the interaction between 19 and hBChE. K_ttnb, binding affinity of the product to the hBChE. (D) Michaelis–Menten curves of hMAO-B with increasing concentrations of 19 were carried out with MMTP as substrate (0.067–1.67 mmol/L). Activity at 0, 0.67, 3.33 and 10.00 μmol/L of 19 is showed as dots, squares, triangles, and stars, respectively.

The inhibition constant of 19 was determined on purified recombinant hMAO-B with direct MMTP assay. GraphPad software was used to fit initial velocity measurements at various inhibitor and MMTP doses to the Michaelis–Menten equation. The best fit (determined yielding R² approximately 1.0) was obtained with competitive inhibition model (Fig. 7) The determined K_i value was in the sub-micromolar range (355.6 ± 72.2 nmol/L), which agrees with the IC₅₀ value.

2.2.5. Multifunctional properties of selected QNs and in vitro BBB permeation

The inhibitory potencies of QNs 15, 18, 19, and 32 caught our eye and deserved further exploration in terms of multifunctionality. These compounds were either potent selective hBChE (15 and 32) or dual hBChE/MAO-B inhibitors (18 and 19). Among these, 19 also inhibited hAChE with nanomolar IC₅₀, and was thus seen as the most promising multiple enzyme inhibitor.

Increased oxidative stress and abnormal brain levels of metallic ions in conjunction with tau/Aβ pathology are the main components of AD pathology⁵⁷. A number of small molecules have been developed as antioxidants and further studied in preclinical and clinical studies of neurodegenerative diseases⁵⁸. In the DPPH scavenging assay⁵⁹, which was used to test the antioxidant capacity of QNs, 18 and 19 reduced DPPH with comparable potencies to resveratrol (EC₅₀ = 50.1 μmol/L). Both QNs are phenol derivatives, which are known antioxidants. On the contrary, 15 and 32 with phenolic group alkylated showed no activity in this assay (Table 3).

Table 3.

In vitro radical-scavenging activities (DPPH assay), Aβ_1–42 aggregation inhibition, and PAMPA–BBB permeability for selected QNs.

QN	Radical-scavenging DPPH assay [EC50 (μmol/L)a]	Inhibition of Aβ1–42 aggregationb	BBB–PAMPA permeability
			Papp (nm/s) (mean ± SD)	−LogPe (log [cm/s])	Permeability predictionc
15	n.a.d	n.i.e	722 ± 27	4.1	High
18	119.2 ± 1.2	n.i.e	437 ± 41	4.4	High
19	126.0 ± 0.1	n.i.e	372f ± 32	4.4	High
32	n.a.d	40.7 ± 2.4%∗	503 ± 28	4.3	High

^aEC₅₀ values are means ± SEM (n = 2, in triplicates).

^bInhibition % at 10 μmol/L ligand and 1.5 μmol/L Aβ_1–42. Inhibition % expressed as mean ± SD (n ≥ 3). One-way ANOVA, post-hoc Bonferroni t-test (SigmaPlot 12.0), in comparison to control (DMSO); ∗P < 0.05.

^cBBB permeability −logP_e < 5.6, high; −logP_e > 6.3, low; 5.6 < −logP_e > 6.3, intermediate.

^dn.a., not active–% DPPH reduced at 100 μmol/L concentration lower than 10%.

^en.i., no inhibition <20%.

^fAt least–final donor concentration below detection limit due to properties of “brain sink” acceptor fluid and high permeability of 19.

The imbalance of redox-active copper and zinc ions as well as the production of reactive oxygen species during redox cycling contribute to the redox stress and neurodegeneration⁶⁰. Therapeutic chelation strategies have been examined for the treatment of AD, with some promising results that yet need validation in vivo⁶¹. 8-Hydroxyquinoline core is known for the metal chelation²⁷, thus we also studied chelating capacity of selected nitrones that were expected to chelate metal ions. The absorption spectra of 18 (λ_{abs, max} = 286 nm) incubated with equimolar quantity of Zn²⁺, Cu²⁺, and Al³⁺ showed characteristic bathochromic shift with new absorption maxima at 309, 310 and 296 nm, respectively (Fig. 8A), whereas addition of Fe²⁺ and Fe³⁺ resulted in hyperchromic shift, i.e., the increase in absorbance at 286 nm, and appearance of secondary maxima at ∼230 nm. No major alterations in the absorption spectra were detected for Ca²⁺ and Mg²⁺ ions. Similar was observed for QN 19 (λ_{abs, max} = 287 nm): i) appearance of maxima at 310, 311 and 296 nm for Zn²⁺, Cu²⁺, and Al³⁺, respectively; ii) hypochromic change in absorbance spectra for 19 in the presence of Fe²⁺ and Fe³⁺, and iii) no major alterations in spectra upon incubation with Ca²⁺ and Mg²⁺ (Fig. 8B). This indicated that both QNs generated corresponding complexes with Zn²⁺, Cu²⁺, Al³⁺, Fe²⁺ and Fe³⁺, whereas they did not chelate Ca²⁺ and Mg²⁺. Copper ion is of particular interest in AD⁶², thus the stoichiometry of the Cu²⁺ complexes with 18 and 19 was determined by Yoe-Jones method⁶³, which showed an interception point at 0.5 indicating 18-Cu²⁺ and 19-Cu²⁺ complexes with a 0.5:1 Cu²⁺/QN molar ratio (Fig. 8C and D).

Figure 8.

Metal chelating properties of 18 and 19. UV–Vis spectra of QNs (A) 18 and (B) 19 (30 μmol/L) alone or in the presence of equimolar amount of metal ions in 20 mmol/L HEPES, 150 mmol/L NaCl, pH = 7.4 at ambient temperature. UV–Vis titration of compounds (C) 18 (30 μmol/L) and (D) 19 with Cu²⁺ in buffer.

Aβ aggregation inhibitors employ an alternative approach that would reduce the neurotoxicity of Aβ oligomers and fibrils by preventing their formation, and are in the clinical trial pipeline of AD drug development⁶⁴. Therefore, the capacity of compounds to diminish Aβ aggregation was determined by the ThT fluorescence assay (Table 3), and only aldehyde 32 inhibited self-induced aggregation of Aβ_1–42, whereas 15, 18, and 19 were inactive.

Further point to consider in the development of CNS-active drugs is the permeation of molecules across BBB, since their molecular targets are located in the CNS. For the purpose of determining the passive permeability of substances across the artificial lipid membrane, the BBB parallel artificial membrane permeability (BBB-PAMPA) test was used (Table 3). High permeability was characteristic for all tested compounds, i.e., 15, 18, 19, and 32. The −logP_e values in Table 3 were similar to those of highly permeable reference drugs lidocaine (4.5) and haloperidol (4.0) and were two orders of magnitude higher than the reference drug with low BBB permeability–theophylline (6.4), while sulphasalazine permeability was not detectable, which validated the integrity of the BBB artificial membranes and confirmed the “high” permeability designation in vitro. Therefore, favorable distribution across the BBB is expected for 15, 18, 19, and 32.

2.2.6. Effects of QNs 18 and 19 on cell viability and neuroprotective activity

With promising in vitro activities at our hands, we progressed with two QNs (i.e., 18 and 19) into cellular assays. Firstly, metabolic activities of SH-SY5Y and HepG2 cell lines–two commonly used cell lines in drug development, were determined, and both 18 and 19 reduced these activities at concentrations higher than 10 μmol/L (Fig. 9A and B). Next, cytotoxicity profile of 19, the most promising multifunctional ligand based on the in vitro studies described above, was determined using 7-aminoactinomycin D (7-AAD), a fluorescent intercalator, on SH-SY5Y and microglial BV2 cells (Fig. 9C and D). Treatment of SH-SY5Y and BV2 cells with 19 resulted in a slight increase of 7-AAD positive (7-AAD^pos) cells for the former (Fig. 9C) and no increase of 7-AAD fluorescence up to 5 μmol/L for the latter cell line (Fig. 9D), whereas at 10 μmol/L concentration, compound 19 reduced cell viability to a significant extent.

Figure 9.

In vitro metabolic activity of HepG2 and SH-SY5Y cells treated with 18 and 19, and cytotoxicity profile of 19. (A) SH-SY5Y and (B) HepG2 cells were treated with 18 and 19 (1–100 μmol/L). The MTS assay was used to determine the level of metabolic activity (relative to DMSO). Data are means ± SEM (n = 2, quadruplicates). (C) SH-SY5Y and (D) BV2 cells treated with compound 19 (0.5–10 μmol/L). The cytotoxicity was determined using 7-AAD. Data are relative % of 7-AAD-positive (7-AAD^pos) cells with respect to control. Data (means ± SEM; n = 2, duplicates). ∗P < 0.05.

Oxidative stress and protein aggregation are two hallmarks of neurodegenerative disorders, and both may cause damage to the CNS. To explore the neuroprotective capacity of compound 19, 6-OHDA- and lipopolysaccharide (LPS)-induced toxic models with SH-SY5Y and BV2 cells were used, respectively. As shown in Fig. 10A, treatment with 6-OHDA increased the percentage of 7-AAD^pos cells and reduced cell viability, while the co-treatment with 19 increased cell viability at 0.5 and 1 μmol/L, whereas this neuroprotective activity was lost at 2.5 and 5 μmol/L. No neuroprotective effects on cell viability were observed in LPS-induced BV2 neuroinflammation cell model (Fig. 10B). On the contrary, at 5 μmol/L, 19 reduced LPS-induced increase in nitrite concentration (Fig. 10C), however this was not translated to increased viability of BV2 microglial cells (Fig. 10B). Altogether, QN 19 showed tolerable cytotoxicity and some neuroprotective effects. Thus, it was worth pursuing further in vivo animal models to determine the pro-cognitive effects, which are crucial for potential translation into additional preclinical studies.

Figure 10.

The effects of 19 on the toxicity generated by 6-OHDA in SH-SY5Y cells and LPS-activated BV2 cells. (A) SH-SY5Y cells treated with 6-OHDA (100 μmol/L). (B) BV2 cells stimulated with LPS (1 μg/mL), both in the absence or presence of 19 (0.5–5 μmol/L). 7-AAD staining was used to determine neuroprotective effects. Data (means ± SEM; n = 2, duplicates) are relative percentages of 7AAD^pos cells normalized to the corresponding control. (C) BV2 cells stimulated with LPS (1 μg/mL) in the absence or presence of 19 (0.5–5 μmol/L). Griess assay was used to determine the nitrite concentration in the culture supernatants. Data (means ± SEM, n = 2, duplicates). ∗P<0.05; ∗∗P<0.1.

2.2.7. In vivo behavioral assays

The dual hBChE/MAO-B inhibitor 19 was studied in a scopolamine-induced acute memory impairment model in mice, which has been used extensively to evaluate the potential therapeutics for AD treatment. The in vivo activity of 19 was compared to rivastigmine, the FDA-approved dual AChE/BChE inhibitor used in the therapy of mild to moderate cases of AD⁶⁵. Scopolamine, a nonselective muscarinic receptor antagonist was used to induce amnesia resembling that observed in AD⁶⁶. Two behavioral tests, namely the novel object recognition task (NOR) and the passive avoidance task (PA) were used to assess the effect of 19 and rivastigmine on memory in mice. A rotarod task was first carried out to rule out drug-induced changes in motor coordination and locomotor activity that might affect behavioral assessment. Intraperitoneally (i.p.) administered dose of 19 (30 mg/kg), the highest dose used in the cognitive tests, did not result in any motor impairments in the rotarod test at 6, 18, and 24 rpm. This was a positive outcome despite nanomolar inhibition of hAChE by 19, which might lead to build-up of ACh in the periphery due to inhibition of AChE in basal ganglia and the parasympathetic autonomic system, and resulting in peripheral adverse effects, such as impaired locomotion⁶⁷. In contrast, rivastigmine (2.5 mg/kg) caused muscle tremor and transient motor coordination deficits shortly after administration.

PA is a fear-motivated avoidance task that relies on an aversive stimulus (a foot shock) to assess contextual long-term memory. As a powerful tool to determine the influence of drugs on memory consolidation, the PA task is continuously used in behavioral pharmacology due to its high-throughput and simplicity⁶⁸. During the acquisition trial, there were no significant group differences in the step-through latency (P > 0.05). In the PA task, repeated-measures ANOVA revealed a substantial prolongation of step-through latency in the retention trial (compared to that of to the acquisition trial) in mice that were administered with the vehicle, indicating operational learning and memory in mice (Fig. 11). In the retention trial, there were also significant differences between vehicle-treated mice and scopolamine-treated control mice (P < 0.0001), between control mice treated with scopolamine and 19-treated (30 mg/kg) memory-impaired mice (P < 0.01) and between control mice treated with scopolamine and rivastigmine-treated (doses of one and 2 mg/kg) (P < 0.001). These findings thus confirm the anti-amnesic effects of 19 in PA task, but only at the maximum dosage examined.

Figure 11.

The effects of 19 (i.p., 1, 10, and 30 mg/kg) and rivastigmine (i.p., 0.5, one and 2.5 mg/kg) on learning and memory measured in the PA task in scopolamine-induced amnesic mice (n = 8–10). Results (mean ± SEM on Day 1 (the acquisition trial) and Day 2 (in the retention trial)). A significant overall treatment effect was observed (F [7130] = 9.070, P < 0.0001). The drug × time interaction and time effect were likewise stastictically significant (F [1130] = 57.97, P < 0.0001, and F [7130] = 7.871, P < 0.0001, respectively). Statistics: ANOVA, Dunnett's post hoc comparison. ∗∗P < 0.01, ∗∗∗P < 0.001 and ∗∗∗∗P < 0.0001.

In contrast to the PA task, the NOR task is based on rodents natural tendency to explore novel objects in the absence of external reinforcement, and measures working memory, attention, and preference for novelty in rodents⁶⁹. A significant overall treatment effect was demonstrated by one-way ANOVA (F [6,61] = 6.388, P < 0.0001). During the recognition phase, compound 19 demonstrated a statistically significant (P < 0.01) preference for the novel object, yet the effect was not dose-dependent (Fig. 12). Importantly, in the NOR task, the antiamnesic efficacy of 19 was similar to that of rivastigmine, which was particularly evident at the 1 mg/kg dose of both test compounds. Overall, as proof-of-concept, we demonstrated that QN 19 improved nonspatial-contextual and recognition memory, and did not affect the motor functions of the mice.

Figure 12.

Effects of 19 (i.p., 1, 10, 30 mg/kg) and rivastigmine (i.p., 0.5, 1, 2.5 mg/kg) on scopolamine-amnesic mice (n = 7–12 per treatment group) in NOR task. Scopolamine administration: s.c. 30 min before the T1 trial and 19 or rivastigmine (i.p., 60 min prior to the T1 trial). Results are reported as mean discrimination index (DI) ± SEM. Statistics: ANOVA, Dunnett's post hoc comparison. ∗∗∗P < 0.001 and ∗∗∗∗P < 0.0001.

2.2.8. In vivo assays on transgenic AD mice

The hippocampus is a crucial structure in the brain for learning and memory⁷⁰. Synaptic dysfunction that begins in vulnerable regions (e.g., neocortex and hippocampus) coincide with Aβ peptide accumulation and tau protein hyperphosphorylation⁷¹. On the basis of Aβ hypothesis, mice models of AD have been created to recapitulate the cognitive changes observed in humans. Even though the correlation between Aβ and cognitive decline is still debated, it is generally agreed that extracellular Aβ is at the core of AD pathogenesis. The double transgenic APPSwe/PS1δE9 mice model⁷². which expresses a mutant human presenilin 1 (PS1δE9) and a chimeric mouse/human amyloid precursor protein (APPswe), is a good animal model that overproduces Aβ and leads to the buildup of Aβ plaques in the hippocampus and cerebral cortex⁷³. The APPSwe/PS1δE9 model characteristically develops a large number of plaques before showing cognitive deficits. Extracellular deposition of Aβ in the brain can be detected at 2.5–3 months of age, a change in long-term potentiation at 3 months of age, and evident dysfunction of learning and memory at 6–8 months, and a striking deposition of Aβ in the hippocampus at 6 months of age⁷⁴. Characteristically, the model replicates cognitive behavioral abnormalities, hippocampal atrophy, neuronal degeneration and reactive astrocytosis⁷⁵. These pathological features make the double transgenic APPSwe/PS1δE9 mice a good model for preclinical drug evaluation⁷⁶.

Encouraged by the results of QN 19 in the scopolamine-induced mouse model of cognitive deficit, we analysed the amount of Aβ plaques in the hippocampus and cerebral cortex of the APPSwe/PS1δE9 transgenic model of AD after chronic treatment with 19. Two treatment groups were formed: 1) female mice aged 5 months previously treated for 2 months, and 2) female and male mice aged 10 months previously treated for 4 months. QN 19 (0.62 mg/kg/day) was administered via subcutaneously implanted ALZET® mini-osmotic pumps, which were replaced every 28 days. Histological staining with Thioflavine S detected amyloid deposits in coronal 35 μm-thick sections of the hipoccampus and cerebral cortex⁷⁷. Chronic treatment with QN 19 (0.62 mg/kg/day) decreased the number of Aβ plaques in both cortex (t-test, P = 0.0138) and hippocampus (t-test, P = 0.0368) of female mice at 5 months of age (Fig. 13). In striking contrast, chronic treatment started later was inefficient in both male and female mice (Supporting Information Figs. S7–S8). Consequently, the findings obtained in vivo imply that therapy with 19 may also have disease-modifying effects when initiated at early, presymptomatic stages of AD.

Figure 13.

Effects of 19 (s.c., 0.62 mg/kg/day, implanted mini-osmotic pump, 2-month treatment) on Aβ load in female APPSwe/PS1δE9 double transgenic mice at 5 months of age. Microphotographs show 35 μm-thick brain coronal sections taken of double transgenic female mice brain, (A) control (n = 3) and (B) 19-treated (n = 4). Thioflavin S fluorescence staining shows Aβ plaques (green structures) detected in cerebral cortex and hippocampus. Scale bars: 500 μm. Plots show the results Aβ plaque number in the (C) cortex, and (D) hippocampus. Data (means ± SEM). ∗P < 0.05.

3. Conclusions

In this work, we studied nineteen QNs as multifunctionl ligands for the therapy of neurodegenerative diseases, and identified compound 19 (Fig. 3) as a dual hBChE and hMAO-B inhibitor. Of particular interest is its relative simplicity, which does not suggest ChE or MAO inhibition at the first glance. The peculiar nanomolar hBChE inhibition is the result of two molecules binding to the active site of hBChE, as shown by enzyme kinetics and resolved 19/hBChE crystals structure. Resolved 19/hMAO-B crystal structure also showed the molecular determinants for nanomolar reversible inhibition (K_i = 355.6 ± 72.2 nmol/L). Moreover, 19 is a potent antioxidant, crosses the BBB, is neuroprotective in cell-based assays, and improves contextual and recognition memory in mouse model of AD with scopolamine-induced acute memory impairment. Chronic treatment with 19 also reduced Aβ load in female APPSwe/PS1δE9 double transgenic mice, suggesting a disease-modifying potential of quinolylnitrones. In summary, we conclude that QN 19 is a potent hBChE and hMAO-B inhibitor that merits further proof-of-concept studies to validate these QNs as a promising class of compounds to alleviate the symptoms and underlying causes of AD and PD.

4. Experimental

4.1. Chemistry

4.1.1. General

Compound purification: flash column chromatography (Merck, Silica Gel 60, 40–63 μm). Reaction monitoring: TLC on Merck's Silica Gel 60 F₂₅₄ plates, visualization: UV light (λ = 254 nm), and vanillin or ninhydrin stains. Following extraction, the organic phases were dried over anh. Mg₂SO₄, filtered and evaporated. Melting points are uncorrected (Reichert Thermo Galen Kofler block). ¹H NMR and 13C NMR were recorded on Bruker Avance 300 (300 MHz), 400 III HD (400 MHz) and 500 II HD (500 MHz) spectrometers, samples dissolved in CDCl₃ or DMSO-d₆. In ¹³C NMR spectra, solvent signal were used as references, whereas TMS was used as the reference in 1H. Chemical shifts (δ) reported in ppm, coupling constants (J) in Hz, signal multiplicities are multiplet (m), singlet (s), doublet (d), triplet (t), quartet (q), septuplet (sept), doublet of doublets (dd), triplet of doublets (td). IR spectra (KBr discs) were recorded on a PerkinElmer Spectrum One B spectrometer. MS were recorded on an Agilent HP 1100 LC/MS spectrometer, and HRMS on an Agilent 6520 Accurate-Mass QTOF LC/MS. Elemental CNH analyses nitrones 1–19, and oxime 33 were recorded on a Carlo Erba EA 1108 (result: ±0.4% of expected).

4.1.2. General procedure (GP) for the synthesis of QNs

GP1: A solution of the corresponding carbaldehyde (1 equiv.), Na₂SO₄ (2 equiv.), NaHCO₃ (1.5 equiv.) and an appropriate N-alkylhydroxylamine hydrochloride (1.5 equiv.) in THF (5 mL) was stirred at 90 °C for 1.5–4.5 h under microwave irradiation (MWI), solvent evaporated and compound purified by column chromatography (CC).

GP2: A solution of the corresponding carbaldehyde (1 equiv.), Na₂SO₄ (3 equiv.), NaOAc (2 equiv.) and an N-alkylhydroxylamine hydrochloride (1.5 equiv.) in EtOH (5 mL) was stirred at 90 °C for 2–3 h under MWI, solvent evaporated and compound purified by CC.

3-(1,3-Dioxolan-2-yl)-8-methoxy-2-(3-(piperidin-1-yl)propoxy)quinoline (21). A solution of 20⁷⁸ (236 mg, 1 equiv.), 3-(piperidin-1-yl)propan-1-ol (1.5 equiv.), t-BuOK (2 equiv.), NaI (1 equiv.) in dioxane (3 mL) was stirred at 110 °C for 6 h (MWI), the solvent evaporated and compound 21 purified by CC (DCM/MeOH, 6%). 21 (white solid, 246.5 mg, 74%): mp 182–184 °C; IR ν 2942, 2666, 1626, 1503, 1433, 1385, 1372, 1108 cm⁻¹; ¹H NMR (400 MHz, CDCl₃) δ 8.14 (s, 1H), 7.37–7.22 (m, 2H), 7.20 (d, J = 2.4 Hz, 1H), 6.98 (d, J = 7.2 Hz, 1H), 6.08 (s, 1H), 4.62 (t, J = 5.4 Hz, 2H), 4.15–3.99 (m, 4H), 3.96 (s, 3H), 3.16–2.64 (m, 5H), 2.37 (d, J = 7.4 Hz, 2H), 1.92 (s, 4H), 1.57 (s, 2H); ¹³C NMR (101 MHz, CDCl₃) δ 158.8, 153.9, 137.7, 136.2, 126.0, 124.5, 122.2, 120.1, 109.5, 99.3, 65.3, 63.3 (2C), 56.3, 55.4 (2C), 53.5, 24.4 (3C), 22.6. HRMS (ESI+) Calcd. for [C₂₁H₂₈N₂O₄]⁺: 373.2122. Found: 373.2103 [M + H]⁺.

6-Methoxy-2-(3-(piperidin-1-yl)propoxy)quinoline-3-carbaldehyde (22). HCl (2 mol/L, 2.4 equiv.) was added dropwise to 21 (231.5 mg, 1 equiv.) in THF (2.5 mL) at rt. After 1 h, DCM was added and washed with sat. NaHCO₃(aq) (3 × 5 mL), and sat. brine. 22 (pale yellow solid, 121.2 mg, 99%): mp 109–111 °C; IR ν 2940, 1692, 1612, 1601, 1267, 1090 cm⁻¹; ¹H NMR (300 MHz, CDCl₃) δ 10.50 (d, J = 1.1 Hz, 1H), 8.57 (d, J = 1.0 Hz, 1H), 7.50–7.34 (m, 2H), 7.13 (dd, J = 7.7, 1.4 Hz, 1H), 4.69 (t, J = 6.4 Hz, 2H), 4.05 (s, 3H), 2.77–2.36 (m, 6H), 2.14 (t, J = 7.6 Hz, 2H), 1.77–1.56 (m, 4H), 1.53–1.36 (m, 2H); 13C NMR (126 MHz, CDCl₃) δ 189.4, 160.5, 153.9, 140.3, 139.8, 125.4, 125.0, 121.6, 120.1, 111.7, 65.1, 56.4, 56.2 (2C), 54.6, 26.3, 25.7 (2C), 24.2. HRMS (ESI+) Calcd. for [C₁₉H₂₄N₂O₃]⁺: 329.1860. Found: 329.1848 [M + H]⁺.

(Z)-N-tert-Butyl-1-(8-methoxy-2-(3-(piperidin-1-yl)propoxy)quinolin-3-yl)methanimine oxide (1). Following GP1, 1 was synthesized from 22 (61 mg, 1 equiv.), Na₂SO₄ (2 equiv.), NaHCO₃ (1.5 equiv.), N-t-BuNH₂OH × HCl (1.5 equiv.) in THF (5 mL), and purified by CC (DCM/MeOH, 5%). 1 (pale yellow solid, 48 mg, 65%): mp 185–187 °C; IR ν 3416, 2961, 1594, 1425, 1346, 1266, 1091 cm⁻¹; ¹H NMR (400 MHz, CDCl₃) δ 10.20 (s, 1H), 8.09 (s, 1H), 7.41–7.36 (m, 1H), 7.33–7.28 (m, 1H), 7.03 (d, J = 7.6 Hz, 1H), 4.70 (t, J = 6.0 Hz, 2H), 4.01 (s, 3H), 3.57 (d, J = 11.9 Hz, 2H), 3.18 (dd, J = 10.1, 5.4 Hz, 2H), 2.70–2.50 (m, 4H), 2.37–2.22 (m, 2H), 1.95–1.80 (m, 4H), 1.65 (s, 9H); ¹³C NMR (101 MHz, CDCl₃) δ 157.4, 153.7, 137.4, 136.7, 126.6, 124.9, 123.6, 121.3, 115.8, 110.14, 71.9, 63.39, 56.2, 55.2, 53.4 (2C), 28.4 (3C), 24.1, 22.6 (2C), 22.2. HRMS (ESI+) Calcd. for [C₂₃H₃₃N₃O₃]⁺: 400.2595. Found: 400.2598 [M + H]⁺.

(Z)-N-Benzyl-1-(8-methoxy-2-(3-(piperidin-1-yl)propoxy)quinolin-3-yl)methanimine oxide (2). Following GP1, 2 was synthesized from 22 (61 mg, 1 equiv.), Na₂SO₄ (2 equiv.), NaHCO₃ (1.5 equiv.), N-BnNH₂OH × HCl (1.5 equiv.) in THF (5 mL), and purified by CC (DCM/MeOH, 4%). 2 (yellow solid 54.7 mg, 58%): mp 104–106 °C; IR ν 3401, 2930, 1594, 1419, 1261, 1092 cm⁻¹; ¹H NMR (400 MHz, CDCl₃) δ 10.08 (s, 1H), 8.22 (s, 1H), 7.71–7.54 (m, 2H), 7.50–7.21 (m, 5H), 7.02 (d, J = 7.7 Hz, 1H), 5.21 (s, 2H), 4.58 (t, J = 6.0 Hz, 2H), 4.00 (s, 3H), 2.89–2.59 (m, 6H), 2.27 (t, J = 7.8 Hz, 2H), 1.92–1.74 (m, 4H), 1.64–1.46 (m, 2H); ¹³C NMR (101 MHz, CDCl₃) δ 157.6, 153.7, 137.6, 137.2, 133.7, 129.5 (2C), 128.9 (2C), 128.5 (2C), 126.3, 124.7, 121.3, 115.6, 110.2, 71.5, 64.1, 56.3 (OCH₃), 55.6, 54.1 (2C), 25.5, 24.5 (2C), 23.4. HRMS (ESI+) Calcd. for [C₂₆H₃₁N₃O₃]⁺: 434.2438. Found: 434.2450 [M + H]⁺.

3-(1,3-Dioxolan-2-yl)-8-methoxy-2-(3-(4-(prop-2-yn-1-yl)piperazin-1-yl)propoxy)quinoline (23). A solution of 20⁷⁸ (236 mg, 1 equiv.), 3-(4-(prop-2-yn-1-yl)piperazin-1-yl)propan-1-ol³⁴ (1.5 equiv.), t-BuOK (2 equiv.), NaI (1 equiv.) in dioxane (3 mL) was stirred at 110 °C for 6 h (MWI), the solvent evaporated and compound 23 isolated by CC (DCM/MeOH, 4%). 23 (yellow solid, 252 mg, 69%): mp 96–98 °C; IR ν 3235, 2929, 1628, 1343, 1265, 1105 cm⁻¹; ¹H NMR (500 MHz, CDCl₃) δ 8.18 (s, 1H), 7.35–7.24 (m, 2H), 7.02 (dd, J = 7.7, 1.4 Hz, 1H), 6.15 (d, J = 0.7 Hz, 1H), 4.63 (t, J = 6.4 Hz, 2H), 4.18–4.04 (m, 4H), 4.02 (s, 3H), 3.29 (d, J = 2.5 Hz, 2H), 2.80–2.39 (m, 10H), 2.24 (t, J = 2.4 Hz, 1H), 2.19–2.03 (m, 2H); ¹³C NMR (126 MHz, CDCl₃) δ 159.3, 153.9, 136.0, 125.8, 124.1, 122.3 (2C), 120.1, 109.4, 99.3, 78.8, 73.2, 65.3, 64.5 (2C), 56.3, 55.3, 53.0 (2C), 51.8 (2C), 46.8, 26.4. HRMS (ESI+) Calcd. for [C₂₃H₂₉N₃O₄]⁺: 412.2231. Found: 412.2221 [M + H]⁺.

8-Methoxy-2-(3-(4-(prop-2-yn-1-yl)piperazin-1-yl)propoxy)quinoline-3-carbaldehyde (24). HCl (2 mol/L, 2.5 equiv.) was added dropwise 23 (253 mg, 1 equiv.) in THF (2.5 mL) at rt. After 1 h, DCM was added and washed with sat. NaHCO₃(aq) (3 × 5 mL), sat. brine, and evaporated. 24 (yellow solid, 222 mg, 99%): mp 147–149 °C; IR ν 2926, 1688, 1602, 1423, 1266, 1093 cm⁻¹; ¹H NMR (500 MHz, CDCl₃) δ 10.50 (d, J = 0.6 Hz, 1H), 8.57 (d, J = 0.6 Hz, 1H), 7.46–7.42 (m, 1H), 7.35 (td, J = 7.9, 0.5 Hz, 1H), 7.14 (dd, J = 7.7, 1.2 Hz, 1H), 4.71 (t, J = 6.4 Hz, 2H), 4.05 (s, 3H), 3.31 (dd, J = 2.5, 0.6 Hz, 2H), 2.92–2.42 (m, 10H), 2.25 (td, J = 2.4, 0.6 Hz, 1H), 2.17–2.07 (m, 2H); ¹³C NMR (126 MHz, CDCl₃) δ 189.3, 160.5, 153.9, 140.3, 139.7, 125.4, 124.9, 121.6, 120.1, 111.7, 78.8, 73.2, 65.0, 56.4, 55.3, 53.1 (2C), 51.9 (2C), 46.8, 26.4. HRMS (ESI+) Calcd. for [C₂₁H₂₅N₃O₃]⁺: 368.1969. Found: 368.1976 [M + H]⁺.

(Z)-N-tert-Butyl-1-(8-methoxy-2-(3-(4-(prop-2-yn-1-yl)piperazin-1-yl)propoxy)quinolin-3-yl)methanimine oxide (3). Following GP1, 3 was synthesized from 24 (75 mg, 1 equiv.), Na₂SO₄ (2.0 equiv.), NaHCO₃ (1.5 equiv.), N-t-BuNH₂OH × HCl (1.5 equiv.), THF (5 mL), and purified by CC (DCM/MeOH, 5%). 3 (white solid, 77.6 mg, 86%): mp 131–133 °C; IR ν 3260, 2808, 1594, 1495, 1350, 1262, 1126 cm⁻¹; ¹H NMR (400 MHz, CDCl₃) δ 10.16 (s, 1H), 8.12 (s, 1H), 7.38 (dd, J = 8.2, 1.3 Hz, 1H), 7.31–7.24 (m, 1H), 7.02 (dd, J = 7.8, 1.2 Hz, 1H), 4.63 (t, J = 6.5 Hz, 2H), 4.02 (s, 3H), 3.36–3.28 (m, 2H), 2.79–2.48 (m, 10H), 2.30–2.23 (m, 1H), 2.17–2.06 (m, 2H), 1.63 (s, 9H); ¹³C NMR (101 MHz, CDCl₃) δ 157.9, 153.7, 137.2, 137.1, 126.4, 124.5, 123.9, 121.3, 115.9, 110.0, 78.7, 73.5, 71.6, 64.7, 56.3, 55.6, 53.14 (2C), 51.6 (2C), 46.8, 28.4 (3C), 26.3. HRMS (ESI+) Calcd. for [C₂₅H₃₄N₄O₃]⁺: 439.2704. Found: 439.2713 [M + H]⁺.

(Z)-N-Benzyl-1-(8-methoxy-2-(3-(4-(prop-2-yn-1-yl)piperazin-1-yl)propoxy)quinolin-3-yl)methanimine oxide (4). Following GP1, 4 was synthesized from 24 (75 mg, 1 equiv.), Na₂SO₄ (1.5 equiv.), NaHCO₃ (2.0 equiv.) and N-BnNH₂OH × HCl (1.5 equiv.), THF (5 mL), and purified by CC (DCM/MeOH, 4%). 4 (solid, 91 mg, 94%): mp 153–155 °C; IR ν 2961, 1596, 1495, 1419, 1264, 1093 cm⁻¹; ¹H NMR (400 MHz, CDCl₃) δ 10.09 (s, 1H), 8.26 (s, 1H), 7.69–7.50 (m, 2H), 7.44–7.23 (m, 5H), 7.09–6.99 (m, 1H), 5.23 (s, 2H), 4.60 (t, J = 5.8 Hz, 2H), 4.00 (s, 3H), 3.34 (d, J = 2.5 Hz, 2H), 3.20–2.64 (m, 10H), 2.42–2.26 (m, 3H); ¹³C NMR (101 MHz, CDCl₃) δ 157.5, 153.7, 137.6, 137.1, 133.7, 129.6 (2C), 128.9 (2C), 128.4, 126.4 (2C), 124.8 (2C), 121.3, 115.6, 110.2, 74.2, 71.5, 63.6, 56.3, 54.8, 52.4 (2C), 49.8 (2C), 46.6, 25.1. HRMS (ESI+) Calcd. for [C₂₈H₃₂N₄O₃]⁺: 473.2547. Found: 473.2559 [M + H]⁺.

2-Chloro-8-hydroxyquinoline-3-carbaldehyde (26). A cooled (−78 °C) solution of 25 (442 mg, 1 equiv.) in DCM (14 mL), was treated with BBr₃ (3 equiv.), and stirred at −78 °C for 30 min and at rt for 3 h. Water (5 mL) was added at 0 °C, and the extracted using EtOAc/H₂O. Organic phases were washed with sat. brine, and evaporated. 26 (310 mg, 75%): mp 163–165 °C; IR ν 3336, 1685, 1678, 1466, 1191 cm⁻¹; ¹H NMR (300 MHz, DMSO-d₆) δ 10.36 (s, 1H), 8.88 (s, 1H), 7.74–7.64 (m, 1H), 7.55 (t, J = 7.7 Hz, 1H), 7.29 (dd, J = 7.7, 1.3 Hz, 1H) (OH signal missing); ¹³C NMR (126 MHz, CDCl₃) δ 188.9, 151.4, 148.6, 140.6, 139.2, 129.4, 126.9, 126.7, 120.0, 114.8. HRMS (ESI+) Calcd. for [C₁₀H₆ClNO₂]⁺: 208.0160. Found: 208.0152 [M + H]⁺.

2-Chloro-8-(3-(piperidin-1-yl)propoxy)quinoline-3-carbaldehyde (27). To a solution of 26 (296 mg, 1 equiv.) in CHCl₃ (5 mL)/H₂O (1 mL), K₂CO₃ (3 equiv.) and 1-(3-chloropropyl)piperidine (1.3 equiv.) were added, and stirred vigorously at 80 °C for 48 h, solvent evaporated and compound 27 purified by CC (DCM/MeOH, 4%). 27 (yellow gum, 217.9 mg, 65%): IR ν 3434, 2930, 1696, 1464, 1366, 1115, 1043 cm⁻¹; ¹H NMR (500 MHz, CDCl₃) δ 10.57 (s, 1H), 8.70 (s, 1H), 7.64–7.49 (m, 2H), 7.29 (dd, J = 6.9, 2.2 Hz, 1H), 4.32 (q, J = 8.0, 7.4 Hz, 2H), 2.60 (t, J = 7.3 Hz, 2H), 2.47 (s, 4H), 2.21 (p, J = 6.9 Hz, 2H), 1.62 (p, J = 5.4 Hz, 4H), 1.50–1.43 (m, 2H); ¹³C NMR (126 MHz, CDCl₃) δ 189.4, 154.0, 149.1, 140.1, 128.4, 127.8 (2C), 126.7, 121.0, 113.3, 68.0, 55.7, 54.5 (2C), 28.3, 26.2 (2C), 24.3. HRMS (ESI+) Calcd. for [C₁₈H₂₁N₂O₂]⁺: 333.1364. Found: 333.1357 [M + H]⁺.

(Z)-N-tert-Butyl-1-(2-chloro-8-(3-(piperidin-1-yl)propoxy)quinolin-3-yl)methanimine oxide (5). Following GP1, 5 was synthesized from 27 (100 mg, 1 equiv.), Na₂SO₄ (2.0 equiv.), NaHCO₃ (1.5 equiv.), N-t-BuNH₂OH × HCl (1.5 equiv.), THF (5 mL), and purified by preparative TLC chromatography (DCM/MeOH, 10%). 5 (white solid, 42 mg, 35%): mp 179–181 °C; IR ν 3406, 2941, 1656, 1568, 1483, 1273, 1117, 1045 cm⁻¹; ¹H NMR (500 MHz, CDCl₃) δ 10.35 (s, 1H), 8.29 (d, J = 0.6 Hz, 1H), 7.56–7.44 (m, 2H), 7.16 (dd, J = 5.1, 3.9 Hz, 1H), 4.32 (t, J = 6.2 Hz, 2H), 2.89 (d, J = 7.3 Hz, 6H), 2.40 (t, J = 7.2 Hz, 2H), 1.91–1.81 (m, 4H), 1.68 (s, 9H), 1.61–1.53 (m, 2H); ¹³C NMR (126 MHz, CDCl₃) δ 153.4, 147.9, 138.6, 137.0, 128.5, 127.8, 125.2, 123.2, 121.2, 112.1, 72.5, 67.5, 55.5, 54.0 (2C), 28.3 (3C), 24.3 (3C), 23.3. HRMS (ESI+) Calcd. for [C₂₂H₃₀ClN₃O₂]⁺: 404.2099. Found: 404.2101 [M + H]⁺.

(Z)-N-Benzyl-1-(2-chloro-8-(3-(piperidin-1-yl)propoxy)quinolin-3-yl)methanimine oxide (6). Following GP1, 6 was synthesized from 27 (100 mg, 1 equiv.), Na₂SO₄ (2 equiv.), NaHCO₃ (1.5 equiv.) and N-BnNH₂OH × HCl (72 mg, 1.5 equiv.), THF (5 mL) at rt, and purified by CC (DCM/MeOH, 5%). 6 (white solid, 130 mg, 99%): mp 207–209 °C; IR ν 3419, 2930, 1486, 1271, 1115, 1081 cm⁻¹; ¹H NMR (500 MHz, CDCl₃) δ 10.25 (s, 1H), 8.12 (s, 1H), 7.62–7.37 (m, 7H), 7.15 (s, 1H), 5.17 (s, 2H), 4.32 (t, J = 5.6 Hz, 2H), 3.26 (d, J = 7.7 Hz, 2H), 3.23–2.92 (m, 4H), 2.56 (t, J = 7.6 Hz, 2H), 2.16–1.98 (m, 4H), 1.76–1.53 (m, 2H); ¹³C NMR (126 MHz, CDCl₃) δ 153.2, 147.3, 138.6, 137.4, 132.6, 129.4, 129.3, 129.2 (2C), 129.1, 128.3 (2C), 127.9, 122.8, 121.6, 112.8, 72.3, 67.2, 55.3, 53.6 (2C), 24.2, 23.0 (2C), 22.4. HRMS (ESI+) Calcd. for [C₂₅H₂₈ClN₃O₂]⁺: 438.1943. Found: 438.1940 [M + H]⁺.

1-(3-Chloropropyl)-4-(prop-2-yn-1-yl)piperazine (28). To a cooled solution (0 °C) of 1-(3-chloropropyl)piperazine × 2HCl (702 mg, 1.5 equiv.) and Et₃N (2 equiv.) in DCM (5 mL), propargyl bromide (3 equiv.) was added, the mixture stirred at rt for 24 h, then treated with sat. NaHCO₃(aq), the organic layer separated, washed with sat. brine, evaporated, and compound 28 purified by CC (DCM/MeOH, 1%–2%). 28 (colorless oil, 312 mg, 52%): ¹H NMR (400 MHz, CDCl₃) δ 3.53 (t, J = 6.6 Hz, 2H), 3.23 (d, J = 2.5 Hz, 2H), 2.54–2.39 (m, 10H), 2.18 (t, J = 2.5 Hz, 1H), 1.88 (p, J = 6.7 Hz, 2H); ¹³C NMR (101 MHz, CDCl₃) δ 78.8, 73.2, 55.4, 53.1 (2C), 51.9 (2C), 46.8, 43.2, 29.9.

2-Chloro-8-(3-(4-(prop-2-yn-1-yl)piperazin-1-yl)propoxy)quinoline-3-carbaldehyde (29). To a solution of 26 (104 mg, 1 equiv.) in CHCl₃/H₂O, 5/1 (10 mL), 28 (1.5 equiv.), K₂CO₃ (3 equiv.), and KI (0.2 equiv.) were added, and the mixture stirred for 24 h at 85 °C, the solvents evaporated, and compound 29 purified (DCM/MeOH 1%–5%). 29 (pale yellow solid, 323 mg, 83%): mp 133–135 °C; IR ν 3120, 2820, 1711, 1460 cm⁻¹; ¹H NMR (400 MHz, CDCl₃) δ 10.55 (s, 1H), 8.69 (s, 1H), 7.53–7.51 (m, 2H), 7.26–7.24 (m, 1H), 4.32 (t, J = 6.6 Hz, 2H), 3.31 (d, J = 2.4 Hz, 2H), 2.55 (m, 10H), 2.25 (t, J = 2.4 Hz, 1H), 2.24–2.21 (m, 2H); ¹³C NMR (101 MHz, CDCl₃) δ 189.5, 154.0, 149.2, 141.5, 140.2, 128.5, 127.9, 126.8, 121.3, 113.5, 78.7, 73.4, 67.8, 55.0 (2C), 53.0 (2C), 46.8, 29.8. HRMS (ESI+) Calcd. for [C₂₀H₂₂ClN₃O₂]⁺: 372.1434. Found: 372.1442 [M + H]⁺.

(Z)-N-tert-Butyl-1-(2-chloro-8-(3-(4-(prop-2-yn-1-yl)piperazin-1-yl)propoxy)quinolin-3-yl)methanimine oxide (7). Following GP2, 7 was synthesized from 29 (96 mg, 1 equiv.), Na₂SO₄ (2 equiv.), NaOAc (1.6 equiv.), N-t-BuNH₂OH × HCl (1.6 equiv) EtOH (7 mL), and purified by CC (DCM/MeOH, 1%–8%).7 (white solid, 59 mg, 52%): mp 183–185 °C; IR ν 3421, 2930, 1496, 1275, 1117 cm⁻¹; ¹H NMR (300 MHz, CDCl₃) δ 10.34 (s, 1H), 8.29 (s, 1H), 7.47–7.45 (m, 2H), 7.15 (dd, J = 6.8, 2.3 Hz, 1H), 4.31 (t, J = 6.5 Hz, 2H), 3.31 (d, J = 2.5 Hz, 2H), 2.73–2.68 (m, 10H), 2.26 (t, J = 2.5 Hz, 1H), 2.27–2.24 (m, 2H), 1.67 (s, 9H); ¹³C NMR (126 MHz, CDCl₃) δ 153.8, 147.9, 138.6, 137.0, 128.5, 127.9, 125.3, 123.2, 120.8, 111.7, 78.6, 73.4, 72.5, 67.7, 59.0, 55.0 (2C), 52.9 (2C), 46.7, 28.3 (3C), 25.3; HRMS (ESI+) Calcd. for [C₂₄H₃₁ClN₄O₂]⁺: 443.2208. Found: 443.2215 [M + H]⁺.

(Z)-N-Benzyl-1-(2-chloro-8-(3-(4-(prop-2-yn-1-yl)piperazin-1-yl)propoxy)quinolin-3-yl)methanimine oxide (8). Following GP2, 8 was synthesized from 29 (82 mg, 1 equiv.), Na₂SO₄ (2 equiv.), NaOAc (1.6 equiv.), N-BnNH₂OH × HCl (1.6 equiv.) and EtOH (7 mL), and purified by CC (DCM/MeOH, 1%–8%). 8 (white solid, 103 mg, 99%): mp 203–206 °C; IR ν 3420, 2928, 1486, 1269, 1113 cm⁻¹; ¹H NMR (300 MHz, CDCl₃) δ 10.23 (s, 1H), 8.13 (s, 1H), 7.53 (m, 2H), 7.45 (m, 4H), 7.44 (m, 1H), 7.15 (dd, J = 6.8, 2.3 Hz, 1H), 5.16 (s, 2H), 4.29 (t, J = 6.7 Hz, 2H), 3.30 (d, J = 2.3 Hz, 2H), 2.65 (m, 10H), 2.25 (t, J = 2.3 Hz, 1H), 2.18 (p, J = 6.7 Hz, 2H); ¹³C NMR (126 MHz, CDCl₃) δ 153.8, 147.3, 138.8, 137.4, 132.8, 129.5, 129.5 (2C), 129.4, 129.2 (2C), 128.1, 127.9, 122.8, 120.8, 111.8, 78.8, 73.4, 72.3, 67.7, 55.0 (2C), 53.1 (2C), 51.8, 46.9, 26.3; HRMS (ESI+) Calcd. for [C₂₇H₂₉ClN₄O₂]⁺: 477.2051. Found: 477.2038 [M + H]⁺.

(Z)-N-tert-Butyl-1-(2-chloro-8-methoxyquinolin-3-yl)methanimine oxide (9). Following GP2, 9 was synthesized from 25 (56 mg, 1 equiv.), Na₂SO₄ (2 equiv.), NaOAc (1.6 equiv.), N-t-BuNH₂OH × HCl (1.5 equiv.) and EtOH (7 mL), and purified by CC (hexane/EtOAc, 9/1). 9 (white solid, 48 mg, 66%): mp 185–188 °C; ¹H NMR (300 MHz, CCCl₃) δ 10.37 (d, J = 0.5 Hz, 1H), 8.31 (d, J = 0.5 Hz, 1H), 7.50–7.48 (m, 2H), 7.12 (dd, J = 6.7, 2.3 Hz, 1H), 4.08 (s, 3H), 1.69 (s, 9H). HRMS (ESI+) Calcd. for [C₁₅H₁₇ClN₂O₂]⁺: 293.1051. Found 293.1061 [M + H]⁺.

(Z)-N-Benzyl-1-(2-chloro-8-methoxyquinolin-3-yl)methanimine oxide (10). Following GP2, 10 was synthesized from 25 (56 mg, 1 equiv.), Na₂SO₄ (71 mg, 2 equiv.), NaOAc (33 mg, 1.6 equiv.), N-BnNH₂OH × HCl (60 mg, 1.5 equiv.) and EtOH (7 mL), and purified by CC (hexane/EtOAc, 9/1). 10 (white solid, 64 mg, 79%): mp 187–189 °C; ¹H NMR (400 MHz, CDCl₃) δ 10.25 (s, 1H), 8.13 (s, 1H), 7.53 (m, 2H), 7.48–7.45 (m, 2H), 7.44–7.33 (m, 3H), 7.10 (dd, J = 7.5, 1.5 Hz, 1H), 5.16 (s, 2H), 4.05 (s, 3H); ¹³C NMR (100 MHz, CDCl₃) δ 154.3, 147.3, 138.4, 137.3, 132.7, 129.4 (2C), 129.3 (2C), 129.1 (2C), 128.1, 127.9, 122.9, 120.6, 110.0, 72.3, 56.1. HRMS (ESI+) Calcd. for [C₁₈H₁₅ClN₂O₂]⁺: 327.0894. Found 327.0903 [M + H]⁺.

(Z)-N-tert-Butyl-1-(2-chloro-8-hydroxyquinolin-3-yl)methanimine oxide (11). Following GP2, 11 was synthesized from 26 (104 mg, 1 equiv.), Na₂SO₄ (2 equiv.), NaOAc (1.6 equiv.), N-t-BuNH₂OH × HCl (1.5 equiv) and EtOH (7 mL), and purified by CC (hexane/EtOAc, 9/1). 11 (white solid, 101 mg, 73%): mp 145–148 °C; ¹H NMR (400 MHz, CDCl₃) δ 10.33 (s, 1H), 8.21 (s, 1H), 7.53 (s, 1H), 7.49–7.47 (m, 1H), 7.34 (dd, J = 7.5, 1.1 Hz, 1H), 7.15 (dd, J = 7.5, 1.3 Hz, 1H), 1.61 (s, 9H); ¹³C NMR (100 MHz, CDCl₃) δ 151.1, 147.3, 137.3, 136.5, 128.7, 127.4, 125.1, 123.4, 119.5, 112.7, 72.6, 28.3 (3C). HRMS (ESI+) Calcd. for [C₁₄H₁₅ClN₂O₂]⁺: 279.0894. Found 279.0891 [M + H]⁺.

(Z)-N-Benzyl-1-(2-chloro-8-hydroxyquinolin-3-yl)methanimine oxide (12). Following GP2, 12 was synthesized from 26 (104 mg, 1 equiv.), Na₂SO₄ (2 equiv.), NaOAc (1.6 equiv.), N-BnNH₂OH × HCl (1.5 equiv.) and EtOH (7 mL), and purified by CC (hexanes/EtOAc, 9/1). 12 (white solid, 126 mg, 81%): 139–142 °C; ¹H NMR (400 MHz, CDCl₃) δ 10.29 (s, 1H), 8.10 (s, 1H), 7.53–7.47 (m, 3H), 7.44–7.38 (m, 5H), 7.22 (dd, J = 7.6, 1.3 Hz, 1H), 5.17 (s, 2H); ¹³C NMR (100 MHz, CDCl₃) δ 151.1, 146.7, 137.5, 136.6, 132.7, 129.4 (3C), 129.1 (3C), 128.8, 127.2, 122.9, 119.5, 113.0, 72.4. HRMS (ESI+) Calcd. for [C₁₇H₁₃ClN₂O₂]⁺: 313.0738. Found 313.0726 [M + H]⁺.

8-(3-(Piperidin-1-yl)propoxy)quinoline-2-carbaldehyde (31). To a solution of 30 (104 mg, 1 equiv.) in CHCl₃ (3.6 mL)/water (0.6 mL), K₂CO₃ (249 mg, 1.8 mmol, 3 equiv.) and 1-(3-chloropropyl)piperidine (1.5 equiv.) were added, stirred vigorously for 24 h at 80 °C, the solvent evaporated and compound 31 purified by CC (DCM/MeOH, 7%). 31 (yellow solid, 159 mg, 89%): mp 44–46 °C; IR ν 2932, 1709, 1462, 1323, 1102 cm⁻¹; ¹H NMR (400 MHz, CDCl3) δ 10.20 (s, 1H), 8.18 (dd, J = 8.5, 0.9 Hz, 1H), 7.97 (d, J = 8.4 Hz, 1H), 7.52 (t, J = 8.0 Hz, 1H), 7.37 (dd, J = 8.3, 1.1 Hz, 1H), 7.12 (dd, J = 7.9, 1.2 Hz, 1H), 4.29 (t, J = 6.8 Hz, 2H), 2.62–2.51 (m, 2H), 2.50–2.30 (m, 4H), 2.25–2.12 (m, 2H), 1.55 (p, J = 5.6 Hz, 4H), 1.39 (q, J = 5.7, 4.3 Hz, 2H); ¹³C NMR (101 MHz, CDCl₃) δ 193.9, 155.5, 151.4, 140.1, 137.2, 131.4, 129.8, 119.5, 117.7, 109.9, 67.9, 55.78, 54.6 (2C), 26.33, 25.8 (2C), 24.3. HRMS (ESI+) Calcd. for [C₁₈H₂₂N₂O₂]⁺: 299.1754. Found: 299.1761 [M + H]⁺.

(Z)-N-tert-Butyl-1-(8-(3-(piperidin-1-yl)propoxy)quinolin-2-yl)methanimine oxide (13). Following GP2, 13 was synthesized from 31 (79 mg, 1 equiv.), Na₂SO₄ (2 equiv.), NaOAc (1.2 equiv.), N-t-BuNH₂OH × HCl (1.2 equiv.) and EtOH (5 mL), and purified by CC (DCM/MeOH, 7%). 13 (yellow solid, 52 mg, 52%): mp 135–137 °C; IR ν 3493, 2942, 1615, 1261, 1096 cm⁻¹; ¹H NMR (400 MHz, CDCl₃) δ 9.27 (d, J = 8.8 Hz, 1H), 8.36–8.13 (m, 1H), 7.53–7.41 (m, 2H), 7.16–7.04 (m, 1H), 4.37 (t, J = 5.7 Hz, 2H), 3.61 (d, J = 11.9 Hz, 2H), 3.43–3.28 (m, 2H), 2.68 (dd, J = 15.8, 6.7 Hz, 4H), 2.32 (d, J = 13.8 Hz, 2H), 2.00–1.79 (m, 4H), 1.67 (s, 9H); ¹³C NMR (101 MHz, CDCl₃) δ 137.5, 132.1, 129.6, 127.6, 127.5, 121.9, 120.4, 119.1, 110.2, 110.0, 72.5.66.7, 55.4, 53.5 (2C), 28.4 (3C), 23.9, 22.7 (2C), 22.2. HRMS (ESI+) Calcd. for [C₂₂H₃₁N₃O₂]⁺: 370.2489. Found: 370.24864 [M + H]⁺.

(Z,E)-N-Benzyl-1-(8-(3-(piperidin-1-yl)propoxy)quinolin-2-yl)methanimine oxide (14). Following GP2, 14 was synthesized from 31 (79 mg, 1 equiv.), Na₂SO₄ (2 equiv.), NaOAc (1.2 equiv.), N-BnNH₂OH × HCl (1.2 equiv.) and EtOH (5 mL), and purified by CC (DCM/MeOH, 7%). 14 mixture of Z and E isomers in a 3.5 : one ratio (yellow solid, 60 mg, 55%): mp 108–110 °C; IR ν 3420, 2935, 1600, 1455, 1105 cm⁻¹; ¹H NMR (400 MHz, CDCl₃) 1H NMR (400 MHz, CDCl₃) δ 9.18 [d, J = 8.8 Hz, 1H, H4, major isomer (MI)], 8.28–8.26 (m, 1H, C₆H₅), 8.19 [d, J = 8.8 Hz, 1H, H4, minor isomer (mI)], 8.17 (d, J = 8.8 Hz, 1H, H3, MI), 8.08 (s, 1H, CH N, MI), 7.79 (s, 1H, CH N, mI), 7.74 (dd, J = 8.4, 1.1 Hz, 1H, H3, mI), 7.55 (dd, J = 7.0, 2.5 Hz, 1H, H5, mI), 7.54 (br d, J = 7.0 Hz, 1H, H5), 7.44–7.38 (m, 6H, H6, C₆H₅), 7.45 (t, J = 7.8 Hz, 1H, H6), 7.52–7.37 (m, 4H, C₆H₅), 7.09 (br d, J = 1.5 Hz, 1H, H7, mI), 7.07 (br d, J = 7.5 Hz, 1H, H7, MI), 5.43 (s, 1H, CH₂C₆H₅, mI), 5.15 (s, 1H, CH₂C₆H_5, MI), 4.31 (t, J = 6.2 Hz, 2H, OCH₂), 3.15–2.65 (m, 6H), 2.43–2.41 (m, 2H), 1.98–1.70 (m, 4 H), 1.65–1.51 (m, 2H). HRMS (ESI+) Calcd. for [C₂₅H₂₉N₃O₂]⁺: 404.2333. Found: 404.2326 [M + H]⁺.

8-(3-(4-(Prop-2-yn-1-yl)piperazin-1-yl)propoxy)quinoline-2-carbaldehyde (32).30 (92 mg, 1 equiv.) was dissolved in CHCl₃ (5 mL) and K₂CO₃ (3 equiv.), 28 (1.3 equiv.) and water (1 mL) were added subsequently, stirred vigorously at 80 °C for 48 h, the solvent evaporated and 32 purified by CC (DCM/MeOH, 4%). 32 (yellow solid, 95.5 mg, 53%): mp 99–101 °C; IR ν 3128, 2829, 1715, 1462, 1320, 1094 cm⁻¹; ¹H NMR (400 MHz, CDCl₃) δ 10.20 (s, 1H), 8.20 (d, J = 8.4 Hz, 1H), 7.98 (d, J = 8.5 Hz, 1H), 7.53 (t, J = 8.5 Hz, 1H), 7.39 (d, J = 8.5 Hz, 1H), 7.13 (d, J = 8.4 Hz, 1H), 4.31 (t, J = 6.5 Hz, 2H), 3.24 (s, 2H), 2.76–2.36 (m, 10H), 2.26–2.13 (m, 3H); ¹³C NMR (101 MHz, CDCl₃) δ 193.8, 155.5, 151.4, 140.1, 137.2, 131.4, 129.8, 119.6, 117.8, 110.0, 78.8, 73.2, 67.7, 55.0, 53.0 (2C), 51.8 (2C), 46.8, 26.3. HRMS (ESI+) Calcd. for [C₂₀H₂₃N₃O₂]⁺: 338.0659. Found: 338.1867 [M + H]⁺.

(Z)-N-tert-Butyl-1-(8-(3-(4-(prop-2-yn-1-yl)piperazin-1-yl)propoxy)quinolin-2-yl)methanimine oxide (15). Following GP2, 15 was synthesized from 32 (69 mg, 1 equiv.), Na₂SO₄ (2 equiv.), NaOAc (1.6 equiv.), N-t-BuNH₂OH × HCl (1.5 equiv.) and EtOH (5 mL), and purified by CC (DCM/MeOH, 3%). 15 (yellow gum, 56 mg, 66%): IR ν 3428, 2819, 1451, 1320, 1155, 1104 cm⁻¹; ¹H NMR (400 Hz, CDCl₃) δ 9.26 (d, J = 8.8 Hz, 1H), 8.19 (d, J = 8.8 Hz, 1H), 8.13 (s, 1H), 7.44–7.38 (m, 2H), 7.08–7.06 (m, 1H), 4.29 (t, J = 6.9 Hz, 2H), 3.30–3.28 (m, 2H), 2.63 (s, 10H), 2.32–2.17 (m, 3H), 1.65 (s, 9H); ¹³C NMR (126 MHz, CDCl₃) δ 154.6, 149.8, 140.2, 136.6, 132.6, 129.5, 127.3, 121.7, 119.5, 109.0, 78.7, 73.4, 72.0, 67.7, 55.0, 53.0 (2C), 51.8 (2C), 46.8, 28.3 (3C), 26.3. HRMS (ESI+) Calcd. for [C₂₄H₃₂N₄O₂]⁺: 409.2598. Found: 409.2599 [M + H]⁺.

(E,Z)-N-Benzyl-1-(8-(3-(4-(prop-2-yn-1-yl)piperazin-1-yl)propoxy)quinolin-2-yl)methanimine oxide (16). Following GP2, 16 was synthesized from 32 (96 mg, 1 equiv.), Na₂SO₄ (2 equiv.), NaHCO₃ (1.5 equiv.), N-BnNH₂OH × HCl (1.5 equiv.) and THF (5 mL), and purified by CC (DCM/MeOH, 3%). 16–mixture of E and Z isomers in a 1.5:1 ratio (yellow gum, 93 mg, 75%): IR ν 3429, 2817, 1457, 1320, 1151, 1104 cm⁻¹; ¹H NMR (500 MHz, CDCl₃) δ 1H NMR (500 MHz, CDCl₃) δ 9.18 (d, J = 8.7 Hz, 1H, H4, mI), 8.26–8.24 (m, 1H, C₆H₅), 8.19 (d, J = 8.7 Hz, 1H, H3, mI), 8.17 (d, J = 8.5 Hz, 1H, H4, MI), 8.06 (s, 1H, CH N, mI), 7.81 (d, J = 8.5 Hz, 1H, H3, MI), 7.75 (s, 1H, CH N, MI), 7.55 (dd, J = 7.8, 1.7 Hz, 1H, H5, MI), 7.45 (t, J = 7.8 Hz, 1H, H6), 7.52–7.37 [m, 4H, C₆H₅; H5 (mI)], 7.12 (dd, J = 7.7, 1.2 Hz, 1H, H7, MI), 7.10 (dd, J = 7.7, 1.3 Hz, 1H, H7, mI), 5.44 (s, 1H, CH₂C₆H₅, MI), 5.14 (s, 1H, CH₂C₆H₅, mI), 4.31 (t, J = 6.4 Hz, 2H, OCH₂), 3.31 (d, J = 2.4 Hz, 2H, CH₂C CH, mI), 3.30 (d, J = 2.4 Hz, 2H, CH₂C CH, MI), 2.75–2.50 (m, 10H), 2.25–2.24 (m, 1H, CH₂C CH), 2.24–2.16 (sept, J = 7.5 Hz, 2H, NCH₂CH₂CH₂O). HRMS (ESI+) Calcd. for [C₂₇H₃₀ClN₄O₂]⁺: 443.2442. Found: 443.2441 [M + H]⁺.

(Z)-1-(8-Hydroxyquinolin-2-yl)-N-methylmethanimine oxide (17)^79,80. Following GP2, 17 was synthesized from 30 (173 mg, 1 equiv.), Na₂SO₄ (3 equiv.), NaOAc (1.6 equiv.) N–MeNH₂OH × HCl (1.5 equiv.) and EtOH (7 mL), and purified by CC (hexane/EtOAc, 9/1). 17 (pale yellow solid, 171 mg, 85%): 149–151 °C; ¹H NMR (300 MHz, CDCl₃) δ 9.16 (d, J = 8.8 Hz, 1H), 8.25 (d, J = 8.8 Hz, 1H), 8.03 (br s, 1H), 7.86 (s, 1H), 7.51–7.46 (m, 1H), 7.36 (dd, J = 7.6, 1.3 Hz, 1H), 7.18 (dd, J = 7.6, 1.3 Hz, 1H), 4.01 (s, 3H); ¹³C NMR (100 MHz, CDCl₃) δ 152.1, 147.7, 138.1, 136.8, 136.4, 128.7, 128.6, 121.4, 117.8, 110.1, 55.1. HRMS (ESI+) Calcd. for [C₁₁H₁₀N₂O₂]⁺: 203.0815. Found 203.0824 [M + H]⁺.

(Z)-N-tert-Butyl-1-(8-hydroxyquinolin-2-yl)methanimine oxide (18). Following GP2, 18 was synthesized from 30 (173 mg, 1 equiv.), Na₂SO₄ (3 equiv.), NaOAc (1.6 equiv.), N-t-BuNH₂OH × HCl (1.5 equiv) and EtOH (7 mL), purified by CC (hexane/EtOAc, 9/1). 18 (pale yellow solid, 159 mg, 92%): 103–104 °C; ¹H NMR (400 MHz, CDCl₃) δ 9.16 (d, J = 8.8 Hz, 1H), 8.16 (d, J = 8.8 Hz, 1H), 8.04 (br s, 1H), 7.97 (s, 1H), 7.41–7.39 (m, 1H), 7.27 (dd, J = 7.6, 1.2 Hz, 1H), 7.09 (dd, J = 7.6, 1.2 Hz, 1H), 1.62 (s, 9H); ¹³C NMR (100 MHz, CDCl₃) δ 152.0, 148.5, 138.1, 136.7, 131.5, 128.5, 128.4, 121.7, 117.8, 110.0, 72.1, 28.4 (3C). HRMS (ESI+) Calcd. for [C₁₄H₁₆N₂O₂]⁺: 245.1284. Found 245.1285 [M + H]⁺.

(Z)-N-Benzyl-1-(8-hydroxyquinolin-2-yl)methanimine oxide (19). Following GP2, 19 was synthesized from 30 (173 mg, 1 equiv.), Na₂SO₄ (3 equiv.), NaOAc (1.6 equiv.), N-BnNH₂OH × HCl (1.5 equiv.) and EtOH (7 mL), and purified by CC (hexane/EtOAc, 9/1). 19 (pale yellow solid, 272 mg, 98%): 110–111 °C; ¹H NMR (400 MHz, CDCl₃) δ 9.15 (d, J = 8.8 Hz, 1H), 8.22 (d, J = 8.8 Hz, 1H), 8.00 (br s, 1H), 7.88 (s, 1H), 7.55–7.51 (m, 2H), 7.44–7.37 (m, 4H), 7.26 (dd, J = 7.6, 1.2 Hz, 1H), 7.08 (dd, J = 7.6, 1.2 Hz, 1H), 5.09 (s, 2H); ¹³C NMR (100 MHz, CDCl₃) δ 152.1, 147.7, 138.1, 136.8, 135.3, 132.7, 129.5 (2C), 129.3, 129.2 (2C), 128.7, 128.6, 121.6, 117.8, 110.1, 72.1. HRMS (ESI+) Calcd. for [C₁₇H₁₄N₂O₂]⁺: 279.1128. Found 2,791,128 [M + H]⁺.

8-Hydroxyquinoline-2-carbaldehyde oxime (33). Following GP2, 33 was synthesized from 30 (173 mg, 1 equiv.), Na₂SO₄ (3 equiv.), NaOAc (1.6 equiv.) and NH₂OH × HCl (1.5 equiv.) and EtOH (7 mL), stirred at 90 °C for 10 min, and purified by CC (hexane/EtOAc, 9/1). 33 (white solid, 176 mg, 94%): mp 176–179 °C; ¹H NMR (300 MHz, CDCl₃) δ 8.38 (d, J = 0.6 Hz, 1H), 8.15 (d, J = 8.6 Hz, 1H), 8.05 (br s, 1H), 7.95 (d, J = 8.6 Hz, 1H), 7.85 (br s, 1H), 7.49 (dd, J = 8.3, 7.6 Hz, 1H), 7.34 (dd, J = 8.3, 1.3 Hz, 1H), 7.20 (dd, J = 7.6, 1.3 Hz, 1H); ¹³C NMR (100 MHz, CDCl₃) δ 152.3, 151.2, 149.6, 137.7, 136.6, 128.6, 128.5, 118.7, 117.8, 110.7. HRMS (ESI+) Calcd. for [C₁₀H₈N₂O₂]⁺: 189.0659. Found 189.0652 [M + H]⁺.

4.2. Biology

4.2.1. Inhibition of cholinesterases

The effect of QNs on ChEs was determined by the method of Ellman³⁸. Compounds were incubated with DTNB (370 μmol/L) and the ChEs (approx. 1–2 nmol/L hBChE or 100–200 pmol/L hAChE) in sodium phosphate buffer (0.1 mol/L, pH = 8.0) at rt for 5 min. Enzymatic reactions were started adding the substrate (500 μmol/L butyrylthiocholine iodide [BTCI] for hBChE or acetylthiocholine iodide [ATCI] for hAChE). DMSO concentration was 1% (v/v). Absorbance (λ = 412 nm) was masured for 2 min (Synergy HT reader). The velocities in the presence (v_i) and absence (v_o) of the QNs were determined, and the inhibition expressed as residual activities (RA = v_i/v_o). The IC₅₀ values were calculated in GraphPad Prism 9.3 software (4-parameter logistic function; GraphPad Software, USA).

The progress curves of the BTCI hydrolysis were measured at 412 nm (ε = 13,800 L/(mol × cm) for the product formed) on PerkinElmer Lambda 45 UV/Vis spectrophotometer, with DTNB (1 mmol/L), BTCI (50 μmol/L) and 19 (1–7.5 nmol/L) in 0.6 mL cuvette at 25 °C in phosphate buffer (25 mmol/L, pH = 7.0), taking into the account a 10 s lag period. K_cat = 814 s⁻¹, K_m = 7.5 μmol/L, and K_ttnb = 38 μmol/L were kept fixed, other velocity constants were fitted with ENZO program⁸¹.

4.2.2. Inhibition monoamine oxidases: IC₅₀ determination

Recombinant microsomal hMAOs, HRP, p-tyramine × HCl and Amplex Red were purchased from Sigma–Aldrich.

Briefly, 100 μL of sodium phosphate (50 mmol/L, pH = 7.4) containing 0.05% [v/v] Triton X-114), QNs, and hMAO-A/B were incubated for 15 min at 37 °C in black 96-well microplates. The reaction was initiated by adding substrate – p-tyramine (1 mmol/L), Amplex Red (200 μmol/L) and HRP (2 U/mL). Fluorescence intensity (530/590 nm) was monitored for 30 min at 37 °C (Synergy HT reader). DMSO was used for control experiments (1%, v/v). To determine the blank value (b), buffer replaced the enzyme. RAs were calculated in accordance with Eq. (1):

RA = (v_i–b)/(v_o–b) (1)

Where v_i is velocity in the presence of the QNs, and v₀ is the velocity in the presence of DMSO. Calculations of IC₅₀ values followed steps outlined in Section 4.2.1.

In order to conduct reversibility assay, hMAO-B was incubated at a final concentration of 100-fold with the QNs and control compounds at 10-fold IC₅₀ at 37 °C. Following and incubation period of 15 min, the mixture was then diluted 100-times into the reaction buffer, which contained Amplex Red, HRP, and p-tyramine. DMSO replaced inhibitor solution in control experiments.

4.2.3. DPPH radical-scavenging potency

DPPH (2,2-diphenyl-1-picrylhydrazyl radical) was dissolved in MeOH (140 μmol/L) and added to methanol solution of the QNs (screening at 100 μmol/L, serial dilution of compounds for EC₅₀ determination) or methanol (negative control) on 96-well microtiter plates in triplicates (Brand microplate, pureGrade, F-bottom). The microtiter plate was incubated at rt for 90 min protected from sunlight, and the absorbance (λ = 517 nm) was measured (Synergy HT reader). DPPH free radical % were calculated as Eq. (2):

DPPH radical (%) = [(A₀–A₁) / A₀] × 100 (2)

where A₀ represent the negative control, and A₁ the test compound. Trolox and sesveratrol were used as the positive controls.

4.2.4. Metal-chelating properties

Chelation was determined in HEPES buffer (20 mmol/L, 150 mmol/L NaCl, pH = 7.4) using 96-well microplates. 30 μmol/L QN solution was treated with equimolar concentrations of Cu²⁺, Zn²⁺, Co²⁺, Mg²⁺, Ca²⁺, Fe²⁺, Fe³⁺, and Al³⁺. Ascorbic acid (1 mmol/L) was added to FeCl₂ stock solution to prevent oxidation. The absorption spectra were measured (Synergy HT reader) after 30 min at rt.

The Cu²⁺ binding stoichiometry was determined by titration of a 30 μmol/L solution of the QNs with increasing concetrations of Cu²⁺ (0–150 μmol/L). The absorbances were recorded at the most responsive wavelength after 30 min incubation at room temperature. The absorbance differences in the presence and absence of Cu²⁺ were plotted against the Cu²⁺/compound molar ratio. Data points at the lowest and greatest Cu²⁺/compound ratios approximated the curves, and the intercepts were determined (Yoe-Jones method)⁶³.

4.2.5. In vitro Aβ_1–42aggregation inhibition – Thioflavin-t (ThT) assay

Briefly, HFIP-pretreated Aβ_1–42 (1.5 μmol/L), the selected QNs and ThT at 10 μmol/L were incubated in 96-well microplate at rt in quadriplicates with shaking for 48 h. The fluorescence (λ = 440/490 nm) was measured every 3 min (Synergy™ H4 reader), and the plateau fluorescence intensities with or without QNs were averaged, and the well's average fluorescence at 0 h was subtracted. The Aβ_1–42 aggregation inhibition is calculated as Eq. (3):

%Inh = (1–F_i / F₀) × 100 (3)

where F_i QN- treated fluorescence, and F₀ is the control fluorescence.

4.2.6. In vitro BBB permeation assay

In vitro permeability through the lipid membrane model of selected QNs was investigated using the BBB-PAMPA assay. Briefly, 100 μmol/L donor solutions of the selected QNs and reference drugs were prepared in Pion's Prisma HT® buffer (pH = 7.4). “Brain sink buffer” (Pion, USA) was used as the acceptor fluid. The PAMPA sandwich was incubated for 4 h at 25 °C before sample analysis. All samples were analyzed with an Agilent 1100 HPLC system with an Xterra MS C18 column (3.5 μm, 100 mm × 4.6 mm) at 50 °C and 1% ammonium phosphate buffer at pH = 2.5 (A) with MeCN (B) as the mobile phase. The percentage of B for isocratic elution: 34%, 19%, 31%, 19%, 14%, 26%, 30% and 6% for 15, 18, 19, 32, lidocaine, sulfasalazine, haloperidol and theophylline, respectively, while the detection wavelengths were 246, 302, 296, 262, 230, 362, 246 and 272 nm, respectively. The negative logarithm of the effective permeability (–logP_e) under sink conditions was calculated using the permeability equation⁸². Permeability was binned based on the reference drugs: CNS+, –logP_e < 5.6, high; CNS–, –logP_e > 6.3, low; intermediate values were designated as intermediate permeability.

4.2.7. MAO-B kinetics and X-ray crystallography

Protein production. For steady-state kinetics analysis and X-ray crystallography, detergent-purified samples of recombinant hMAO-B expressed in Pichia pastoris were used⁸³. Purified hMAO-B was stored in potassium phosphate buffer (50 mmol/L, pH = 7.5), 20% glycerol and 0.8% (w/v) β-octylglucoside. NanoDrop ND-1000 was used to measure flavin cofactor absorbance peak at 456 nm, to determine hMAO-B concetration (ε₄₅₆ = 12,000 L/(mol×cm)). Direct spectrophotometric MMTP assay was used for kinetic experiments⁸⁴. Assays were performed in quartz 100 μL cuvettes at 25 °C with a Cary 100 UV/Vis spectrophotometer (Agilent, CA, USA) in HEPES buffer (50 mmol/L, pH = 7.5) containing 0.25% (v/v) reduced Triton X-100. The measurement was started by adding hMAO-B (0.07 μmol/L). The oxidation rate of MMTP substrate (ε₄₂₀ = 25,000 L/(mol×cm)) was monitored over time at different inhibitor 19 concentrations in the 0–10 μmol/L range. The inhibition constant value was determined with non-linear regression by fitting the data with competitive inhibition model using GraphPad 5.0 (GraphPad Software).

X-ray crystallography. Established previously reported protocols were used to determine the 19/hMAO-B structure (Supporting Information Table S2, and Fig. 5). Briefly, hMAO-B (about 50 μmol/L) was gel-filtered in 25 mmol/L potassium phosphate buffer (pH = 7.2) containing 8.5 mmol/L Zwittergent 3–12, and co-crystallized with QN 19 (1 mmol/L) by the sitting-drop method. Crystals were obtained in about 1 week, then transferred into a mother liquor containing glycerol (18%, v/v) and then flash-cooled in liquid nitrogen. The X-ray data were collected at 100 K (beamline ID30A-1/MASSIF-1 ESRF, Grenoble, France). The data processing and scaling (Supporting Information Table S2) were accomplished using XDS⁸⁵ and the CCP4 package⁸⁶. The coordinates of the hMAO-B/safinamide complex⁸⁷ devoid of all water and ligands were used. Electron density inspection and model building was done using Coot⁸⁸ programme. Refinement was performed with the REFMAC5⁸⁹ programme, and Figures prepared in Pymol programme. The structure of hMAO-B/19 complex was deposited into the PDB (ID 7ZW3).

4.2.8. hBChE X-ray crystallography

Crystallization. hBChE was prepared using CHO cells⁹⁰, and purified as previously described⁹¹. Briefly, BChE-specific affinity chromatographic step (Hupresin®) was followed by SEC (Superdex 200). Crystals were obtained at 293 K by the hanging drop method using crystallization buffer MES (100 mmol/L, pH = 6.5), 2.15 mol/L (NH₄)₂SO₄. Metanolic solutions of QNs 15, 19 and 32 (0.1 mol/L) were used for soaking crystals in buffer containing about 2.5 mmol/L of the respective QN. Crystals were cryo-protected in a MES solution (100 mmol/L, pH = 6.5, 2.15 mol/L (NH₄)₂SO₄ and 20% glycerol), before cooling into N₂(l).

Crystallization. X-ray data were collected at ID23-1 and ID30B beamlines of the ESRF (Grenoble, France) at 100 K. Images were either processed with the xia 2/DIALS pipeline from the CCP4 software suite⁹² for 19, the EDNA auto process⁹³ for 15 and the autoPROC⁹⁴ auto process for 32. Phases were attained by molecular replacement (hBChE structure, PDB code 1P0I) with the Phaser program of the Phenix software suite⁹⁵. The hBChE model was refined using successive Phenix. refine and model building (Coot) iterative cycles⁹⁶. Ligand geometry restraints were prepared with Phenix eLBOW⁹⁷ (semi-empirical quantum mechanical method AM1). hBChE structures in complex with QNs 19, 15 and 32 were deposited into the PDB (7QBQ, 7QBR and 7 ZPB, respectively).

4.2.9. Molecular dynamics

The 19-hMAO-B structure was prepared with Schrödinger's Protein Preparation Wizard tool^98,99 (Schrödinger Small Molecule Discovery Suite Release 2021–1). All other hets (except the co-crystallized ligand, crystal waters, FAD), ions, and protein chain B were removed. MD parameters are given in Table 4.

Table 4.

MD parameters.

Software	Desmond/Maestro non-commercial distribution (2020.4)
System builder parameters:
System shape and size	Orthorhombic box (10 Å buffer)
Water model	TIP4P
Electrolyte	150 mmol/L NaCl
Simulation parameters:
Relaxation protocol	desmond_npt_relax.msj (default)
Simulation time	500 ns
Number of frames	5000
Statistical mechanical ensemble	NPT
Temperature	300 K
Pressure	1013 mbar
Thermostat	Nose–Hoover chain (1 ps)
Barostat	Martyna-Tobias-Klein (2 ps)
Timestep	2 fs, RESPA integrator
Cutoff	9.0 Å
Forcefield	OPLS_2005
Analysis	Built-in Desmond tools

4.2.10. Cell-based assays

SH-SY5Y and HepG2 cells were from ATCC, and BV2 cells were a gift from Dr. Minelli. Cell cultures were grown in Advanced Dulbecco's modified Eagle's medium with 10% FBS, l-glutamine (2 mmol/L), streptomycin (50 μg/mL) and penicillin (50 U/mL) in a humidified atmosphere of air/CO₂ (95%/5%) at 37 °C.

Before treating the cells, full medium was exchanged with serum-free medium. QNs 18 and 19 (50 mmol/L stock in DMSO) were used at 0.5–100 μmol/L. For evaluation of neuroprotective effects of compound 19, cells were pretreated with QNs for 1 h, followed by cytotoxic stimuli. SH-SY5Y cells were treated with 100 μmol/L 6-OHDA (Sigma), and BV2 cells were treated with LPS (1 μg/mL, Escherichia coli 055:B5) for 24 h.

MTS assay. SH-SY5Y and HepG2 cells were seeded in 96-well plates (2 × 10⁴/well) and their response to QNs (1–100 μmol/L) was determined using MTS assay. Cells were treated with QNs in serum-reduced medium, and metabolic activity was measured 24 h later using the CellTiter 96® AQ assay (Promega). Absorbance (λ = 492 nm) was measured with Tecan Safire² reader and results reported as a percentage of the control (DMSO).

Cytotoxicity assay. The neuroprotective impact of QNds against cytotoxic effects of neurotoxin 6-OHDA and proinflammatory stimuli of LPS were determined by flow cytometry using fluorescent intercalator 7-aminoactinomycin D (7-AAD). SH-Y5Y (5 × 10⁴/well) and BV2 (5 × 10⁴/well) cells were transferred in 24-well and 12-well culture plates, respectively, and treated next day with 6-OHDA or LPS as described above. After 24 h treatment, cells were collected, washed in cold phosphate-buffered saline (PBS), and labelled with 7-AAD (2 μg/mL) at rt for 10 min, and analyzed by flow cytometry (Attune NxT flow cytometer). The propertion of 7-AAD positive (7-AAD^pos) cells was determined using FlowJo software (FlowJo) and reported compared to control treated cells.

Quantification of nitrite. BV2 cells were transferred in 12-well culture plates (5 × 10⁴/well) and next day stumilated with LPS. After 24 h, cell supernatants were collected by centrifugation (5 min at 1200 rpm), and used for determination of nitrite content as an indicator of nitric oxide (NO) production using Griess reagent kit (Promega). Absorbance (λ = 550 nm) was measured using Tecan Safire² reader. NaNO₂ was utilized as the external standard.

Statistical analyses. Results are presented as means ± SEM (n = 2, in quadruplicate). Statistical analysis: one-way ANOVA, Tukey's post hoc test (GraphPad Prism 6.0, GraphPad Software).

4.2.11. In vivo evaluation

In vivo evaluation Male Albino Swiss (CD-1) mice weighing 18–22 g were obtained from the Faculty of Pharmacy's animal facility (Jagiellonian University Medical College). The animals were kept in cages (10 per cage) at a 22 ± 2 °C, humidity of 55 ± 10%, and a 12 h:12 h light–dark cycle. Ample amounts of food and water were always available to the mice before the start of the experiments. Throughout the experimental period, the prescribed housing conditions for the mice were ensured, including tree litter (Transwior, Poland) and cage enrichment (tunnels, nesting material, wooden igloos, etc.). Mice were randomly selected for behavioral tests, and each treatment group consisted of 7–10 mice. The behavioural assays were conducted between 9 am and 3 pm. Following the in vivo assays, all animals were euthanized immediately. First Local Ethics Committee in Krakow (Permit No. 524/2021) approved the maintenance and treating procedures, and the treatment of the animals fully complied with high ethical standards set forth in relevant EU regulations (Directive 2010/63/EU).

To induce learning and memory deficits, scopolamine hydrobromide (Sigma) was used (1 mg/kg)^100,101. For in vivo testing, scopolamine was prepared in distilled water (Polfa Kutno, Poland), and administered subcutaneously (s.c.) 30 min prior to the behavioral testing during the acquisition (training) phases of NOR/PA tasks. Compound 19 was dissolved in DMSO (3%, v/v) and distilled water and administered i.p. 1 h prior to the acquisition trial of PA or NOR tasks. The procognitive activity of 19 was asessed at three doses: 1, 10 and 30 mg/kg. In the rotarod test, 19 was injected i.p. 60 min before testing. Rivastigmine (rivastigmine tartrate) was purchased from Sigma, and administered i.p. (0.5, 1, and 2.5 mg/kg) 1 h prior to the acquisition trial of NOR/PA tasks.

Behavioral tests—passive avoidance (PA) task. To test the impact of 19 on fear-motivated memory and learning in mice, the PA test was performed according to the previously described protocol¹⁰¹. The PA task is divided into two trials (i.e., acquisition and retention trials) separated by one day. During the conditioning phase, the test compounds (scopolamine, 19, rivastigmine and vehicle) were injected. Treated mice were placed in the white section with the gate closed, and habituation period of 30 s began. Thereafter, the gate was opened, and the mice were followed for 180 s. When the mice entered the black section, the gate was closed and an electric shock (intensity: 0.2 mA, duration: 2 s) was administered. The latency between the opening of the gate and the entry of the mice into the black section was determined. In the retention, drug-off trial, the mice were placed in the white section for a habituation period (30 s), and the latency to enter the black section was measured (the “step-through latency”).

Behavioral tests—novel object recognition task (NOR) task. The protocol used to perform the NOR task was adapted from the literature with some minor modifications^102,103. NOR was performed in opaque black boxes measuring 50 cm × 50 cm × 50 cm. The protocol consisted of 10-min familirisation period (without objects) for 10 min on Day 1 (T₀), a training trial on Day 2 (T₁) and a testing trial on Day 3 (T₂). During T₁ (training trial), two identical objects, i.e., objects A and B, were positioned in the distal corners of the arena (5 cm from the walls). During T₂ (recognition trial), object B was replaced by a new object C. Trials T₁ and T₂ lasted 10 min each. White round plastic containers (height 5 cm, diameter 5 cm) attached to the floor with double-sided tape were used as objects A and B, and green cuboids of similar height were used as objects C. Each mouse was given a different presentation order, and the order in which the objects were placed was also randomized. The animals investigated the objects by looking at them, licking them, sniffing them, or touching them while they were sniffing; however, they did not investigate the objects when they leaned on them, stood on them, or sat on them. Mice that investigated two different objects for a combined total time of less than 5 s at either T₁ or T₂ were not included in the analysis. The formula for calculating the discrimination index (DI) was as Eq. (4):

DI (%)= (EC − EA)/(EA + EC) × 100 (4)

This calculation was based on the exploration time (E) that was spent on both of the objects during the second trial (T₂). Values that were near to zero indicated no preference, whilist values that were positive indicated a preference for exploring unfamilliar objects and negative values indicated preference for exploring known objects.

Rotarod test. The animals were trained for a total of three days on the rotarod apparatus (rod diameter, 2 cm; May Commat RR0711), which rotated at a rate of 18 revolutions per minute, before the actual rotarod test was performed (rpm). During each training session, the mice were placed on the rod for 3 min. Twenty-four hours after the last training session, vehicle or 19 were administered i.p., and the performance of mice in rotarod test was assessed after 1 h (6, 18, and 24 rpm). The inability to maintain balance on the rotarod apparatus for 60 s was defined as evidence of motor impairment.

Data analysis. Statistical analysis: one-way ANOVA, Dunnett's post hoc test, or Student's t-test (GraphPad Prism). A value of P < 0.05 was considered significant.

In vivo evaluation—transgenic AD mice model. Experiments were preformed on adult double transgenic male and female APPswe-PS1δA9 mice. The experiments complied with guidelines for animal welfare and experimentation according to the European Union legislation (Directives 2010/63/UE, ECC/566/2015), and the Spanish Department of Health legislation (R.D. 53/2013). Special efforts were made to minimize the number of mice used. All procedures were approved by the Animal Welfare Committee of the Cajal Institute, by the Institutional Animal Ethics Committee of the Spanish Council for Scientific Research (CSIC), and the experimental animal health authority of the Community of Madrid (PROEX 116.6/21).

Experiments were performed on 10-month-old males and females (n = 14) and 5-month-old females (n = 7), which were binned into two groups – control (n = 10) and treated (n = 11). Two subgroups of mice formed the treated group: the first received treatment for two months (until 5 months of age), and the second was treated for 4 months (until 10 months of age). Mice were treated with 19 (0.62 mg/kg/day) via mini-osmotic pumps (ALZET, model 2004) or vehicle (control group). The pumps were exchanged every 28 days. The osmotic mini-pumps are widely used for continuous drug administration in the study of neurodegenerative diseases. During treatment, body weight was recorded weekly. After treatment, rodents were sacrificed and histological studies were performed.

Tioflavin-S staining and statictical analysis. Reduction of Aβ plaque load in vivo was assessed with the fluorescent dye Thioflavin S, as previously reported¹⁰⁴. Deeply anesthetized animals (chloral hydrate and pentobarbital) were exanguinated by transcardiac perfusion with PBS for 10 s, followed by fixation with 100 mL of a 4% solution of paraformaldehyde (w/v) in phosphate buffer (0.1 mol/L, pH = 7.4, PB). After dissection, brains were postfixed in the fixative solution for 4 h at rt, and cryopreserved in sucrose (30%) in PB at 4 °C 35 μm-thick coronal sections were obtained from the fixed and cryopreserved brains using Leica CM 1950 cryostat. Aβ deposits in the histological sections were detected using a staining procedure with 0.05% thioflavin-S in 50% ethanol. Sections were selected between the levels of Bregma AP –1.28 and −1.64 mm according to a mouse brain atlas. For each animal, ten sections were processed, visualized, and photographed with a Leica DMI8 microscope (Leica DFC 9000 GT camera) using Thunder technology (Leica®). ImageJ software was used to quantify the number of amyloid plaques in the hippocampus and cortex from the microscopic images. Results are reported as means ± SEM. Statistical anaylsis was performed using GraphPad Prism 5.0. Data were compared with an unpaired t-test, and statistical differences are accounted fors at P < 0.05.

Author contributions

Damijan Knez, carried out in vitro assays of nitrones and wrote the article; Daniel Diez-Iriepa carried out the synthesis of nitrones; Mourad Chioua, carried out the synthesis of the nitrones; Andrea Gottinger, carried out kinetics experiments and crystallization studies on hMAO-B; Milica Denic, Fabien Chantegreil, Florian Nachon, and Xavier Brazzolotto, performed the crystallization studies on hBChE; Anna Skrzypczak-Wiercioch and Kinga Sałat, performed in vivo experiments; Anže Meden, performed the molecular dynamics; Anja Pišlar and Janko Kos performed cell-based assays; Simon Žakelj, performed BBB-PAMPA experiment; Jure Stojan, carried out kinetics experiments on hBChE; Julia Serrano, Ana Patricia Fernández, Aitana Sánchez-García, and Ricardo Martínez-Murillo,performed the in vivo assays on double transgenic mice; Claudia Binda, supervised kinetics experiments and crystallization studies on hMAO-B; Francisco López-Muñoz, supervised the synthesis of nitrones; Stanislav Gobec and José; Marco-Contelles edited the article, and supervised the study. All authors have contributed to the analysis of the data, writing of the manuscript and approved the final version of the article.

Conflicts of interest

The authors declare no conflicts of interest.

Appendix A. Supplementary data

The following is the Supplementary data to this article.