The Structural Combination of SIL and MODAG Scaffolds Fails to Enhance Binding to α-Synuclein but Reveals Promising Affinity to Amyloid β

A technique to image α-synuclein (αSYN) fibrils in vivo is an unmet scientific and clinical need that would represent a transformative tool in the understanding, diagnosis, and treatment of various neurodegenerative diseases. Several classes of compounds have shown promising results as potential PET tracers, but no candidate has yet exhibited the affinity and selectivity required to reach clinical application. We hypothesized that the application of the rational drug design technique of molecular hybridization to two promising lead scaffolds could enhance the binding to αSYN up to the fulfillment of those requirements. By combining the structures of SIL and MODAG tracers, we developed a library of diarylpyrazoles (DAPs). In vitro evaluation through competition assays against [3H]SIL26 and [3H]MODAG−001 showed the novel hybrid scaffold to have preferential binding affinity for amyloid β (Aβ) over αSYN fibrils. A ring-opening modification on the phenothiazine building block to produce analogs with increased three-dimensional flexibility did not result in an improved αSYN binding but a complete loss of competition, as well as a significant reduction in Aβ affinity. The combination of the phenothiazine and the 3,5-diphenylpyrazole scaffolds into DAP hybrids did not generate an enhanced αSYN PET tracer lead compound. Instead, these efforts identified a scaffold for promising Aβ ligands that may be relevant to the treatment and monitoring of Alzheimer’s disease (AD).


Introduction
Neurodegenerative diseases such as Parkinson's disease (PD), multiple system atrophy (MSA), and dementia with Lewy bodies (DLB) share the accumulation of α-synuclein (αSYN) fibrils as a hallmark of pathogenesis [1,2]. Although a cell-to-cell prion-like spreading of αSYN oligomers has been widely suggested [3,4], the protein's role in disease progression is still not fully understood [5,6]. A technique to image αSYN fibrils in vivo would  [19] was schematically illustrated.

Development of a DAP Library by the Combination of SIL and MODAG Scaffolds
Two sets of diarylpyrazole (DAP) hybrid compounds were designed and developed based on the combination of the parent scaffolds ( Figure 1) to produce a new class of compounds with promising αSYN binding. Bromo-substituted SIL5 (3) and SIL26 (5) analogs were synthesized by the NiCl2-catalyzed reduction of the nitro group into a primary amine followed by bromination, using t-BuONO and CuBr2 (Scheme 1) [25,26]. The 3bromo phenothiazines were reacted in Pd(PPh3)4-catalyzed cross-couplings with diversely substituted THP-protected 3-aryl stannylpyrazoles followed by acidic deprotection to produce a small library of novel derivatives (Schemes 2 and 3). Their binding properties were evaluated through competition binding assays against [ 3 H]SIL26 and [ 3 H]MODAG−001 on human recombinant αSYN fibrils and synthetic Aβ1-42 fibrils. The in vitro results reported in Table 1 show that DAP hybrids have a significantly lower affinity towards αSYN than both parent scaffolds, with a minimum of 6-fold reduction compared to SIL26 and a 117-to 848-fold decrease compared to MODAG−001 [11,18]. These findings disprove our hypothesis of enhanced binding properties resulting from the combination of pharmacophores from SIL and MODAG structures. as an attempt to achieve interaction with a larger portion of the binding site. αSYN structure described by solid-state NMR [24] (PDB ID: 2N0A) was visualized by Chem3D 20.1 (PerkinElmer Informatics, Waltham, MA, USA) and the ligands' interaction with the binding site 3/13 [19] was schematically illustrated.

Development of a DAP Library by the Combination of SIL and MODAG Scaffolds
Two sets of diarylpyrazole (DAP) hybrid compounds were designed and developed based on the combination of the parent scaffolds ( Figure 1) to produce a new class of compounds with promising αSYN binding. Bromo-substituted SIL5 (3) and SIL26 (5) analogs were synthesized by the NiCl 2 -catalyzed reduction of the nitro group into a primary amine followed by bromination, using t-BuONO and CuBr 2 (Scheme 1) [25,26]. The 3-bromo phenothiazines were reacted in Pd(PPh 3 ) 4 -catalyzed cross-couplings with diversely substituted THP-protected 3-aryl stannylpyrazoles followed by acidic deprotection to produce a small library of novel derivatives (Schemes 2 and 3). Their binding properties were evaluated through competition binding assays against [ 3 H]SIL26 and [ 3 H]MODAG−001 on human recombinant αSYN fibrils and synthetic Aβ 1-42 fibrils. The in vitro results reported in Table 1 show that DAP hybrids have a significantly lower affinity towards αSYN than both parent scaffolds, with a minimum of 6-fold reduction compared to SIL26 and a 117-to 848-fold decrease compared to MODAG−001 [11,18]. These findings disprove our hypothesis of enhanced binding properties resulting from the combination of pharmacophores from SIL and MODAG structures. SAR studies carried out by Wagner et al. on the first set of 3,5-diphenylpyrazoles demonstrated a loss of activity when the bromine substitution is transferred from the meta to the ortho position, highlighting the importance of a planar ligand conformation [13]. We speculate that the constrained flat structure of phenothiazines hampers a successful interaction with the binding site by interrupting the linearity of the two phenyl rings bound to the pyrazole.  SAR studies carried out by Wagner et al. on the first set of 3,5-diphenylpyrazoles demonstrated a loss of activity when the bromine substitution is transferred from the meta to the ortho position, highlighting the importance of a planar ligand conformation [13]. We speculate that the constrained flat structure of phenothiazines hampers a successful interaction with the binding site by interrupting the linearity of the two phenyl rings bound to the pyrazole.
Interestingly, the binding affinity of DAP hybrids to Aβ fibrils was in the range of 1 to 10 nM for most analogs (Table 1), 1 to 2 orders of magnitude lower than the parent scaffold SIL26. Their low Ki values alone are not sufficient to make our hybrid compounds attractive as potential Aβ PET tracers as they are comparable to the FDA-approved compounds (Florbetaben: Ki = 6.70 ± 0.30 nM, Florbetapir: Kd = 3.72 ± 0.30 nM, Flutemetamol: Ki = 0.74 ± 0.38 nM) [27,28]. An Aβ-focused assay comparing DAPs with the abovementioned tracers would be required to validate the correlation. However, our hybrid compounds represent a novel scaffold whose theranostic profile is worth further investigation. In fact, a library of rhodanine-substituted phenothiazines developed by Dao et al. exhibited inhibition of Aβ1-42 aggregation with IC50 values as low as 0.67 ± 0.02 µM and also induced the disaggregation of preformed Aβ1-42 fibrils (IC50 = 0.82 ± 0.10 µM) [29]. As DAP's overall structure and predicted conformation ( Figure 2) partially overlap with these fluorescent probes, we speculated that they could also share some of their physiological properties. An additional study specifically investigating these traits will be necessary in order to confirm our hypothesis. Interestingly, the binding affinity of DAP hybrids to Aβ fibrils was in the range of 1 to 10 nM for most analogs (Table 1), 1 to 2 orders of magnitude lower than the parent scaffold SIL26. Their low K i values alone are not sufficient to make our hybrid compounds attractive as potential Aβ PET tracers as they are comparable to the FDA-approved compounds (Florbetaben: K i = 6.70 ± 0.30 nM, Florbetapir: K d = 3.72 ± 0.30 nM, Flutemetamol: K i = 0.74 ± 0.38 nM) [27,28]. An Aβ-focused assay comparing DAPs with the abovementioned tracers would be required to validate the correlation. However, our hybrid compounds represent a novel scaffold whose theranostic profile is worth further investigation. In fact, a library of rhodanine-substituted phenothiazines developed by Dao et al. exhibited inhibition of Aβ 1-42 aggregation with IC 50 values as low as 0.67 ± 0.02 µM and also induced the disaggregation of preformed Aβ 1-42 fibrils (IC 50 = 0.82 ± 0.10 µM) [29]. As DAP's overall structure and predicted conformation ( Figure 2) partially overlap with these fluorescent probes, we speculated that they could also share some of their physiological properties. An additional study specifically investigating these traits will be necessary in order to confirm our hypothesis.

Enhancement of Structural Flexibility in Ring-Opened DAP Analogs
Based on the hypothesis that the constrained three-dimensional conformation of the phenothiazine ring prevents a successful interaction with the binding site, more flexible DAP hybrid compounds were synthesized to afford a set of analogs that would adjust more effectively to the binding site.
The NH4OAc-catalyzed bromination procedure proposed by Das et al. was modified

Enhancement of Structural Flexibility in Ring-Opened DAP Analogs
Based on the hypothesis that the constrained three-dimensional conformation of the phenothiazine ring prevents a successful interaction with the binding site, more flexible DAP hybrid compounds were synthesized to afford a set of analogs that would adjust more effectively to the binding site.
The NH 4 OAc-catalyzed bromination procedure proposed by Das et al. was modified into a one-pot, sequential substitution of aniline with N-bromosuccinimide and N-iodosuccinimide to generate 2-iodo-4-bromoaniline (7) [30]. The intermediate was further reacted with 3-methoxythiophenol, followed by N-acetylation, to produce 9, the ringopened analog of the phenothiazine-based compound 3 (Scheme 1). As in the previously established synthetic pathway, Pd(PPh 3 ) 4 -catalyzed cross-coupling and acidic deprotection afforded a selection of ring-opened DAP hybrid compounds (Scheme 3).
The minor structural alteration significantly changed the molecule's three-dimensional architecture, conferring more conformational freedom and potentially favoring the binding affinity. Energy minimization of DAP1a and DAP3a structures revealed a discontinuance of the planarity in the diphenyl thioether moiety of the DAP3a most stable conformation ( Figure 2). Together with an electronic change arising from the replacement of a heteroaromatic amine with an aniline, this transformation led to the formation of molecules with very different properties. However, in in vitro assays, all compounds showed a complete loss of competition for αSYN fibrils when tested against both [ 3 H]SIL26 and [ 3 H]MODAG−001 (Table 2), invalidating our hypothesis on flexibility. Additionally, a significant decrease in affinity to Aβ was also observed, highlighting the criticality of three-dimensional conformation in designing new ligands to differentiate between misfolded proteins.

Enhancement of Structural Flexibility in Ring-Opened DAP Analogs
Based on the hypothesis that the constrained three-dimensional conformation of the phenothiazine ring prevents a successful interaction with the binding site, more flexible DAP hybrid compounds were synthesized to afford a set of analogs that would adjust more effectively to the binding site.
The NH4OAc-catalyzed bromination procedure proposed by Das et al. was modified into a one-pot, sequential substitution of aniline with N-bromosuccinimide and Niodosuccinimide to generate 2-iodo-4-bromoaniline (7) [30]. The intermediate was further reacted with 3-methoxythiophenol, followed by N-acetylation, to produce 9, the ringopened analog of the phenothiazine-based compound 3 (Scheme 1). As in the previously established synthetic pathway, Pd(PPh3)4-catalyzed cross-coupling and acidic deprotection afforded a selection of ring-opened DAP hybrid compounds (Scheme 3).
The minor structural alteration significantly changed the molecule's three-dimensional architecture, conferring more conformational freedom and potentially favoring the binding affinity. Energy minimization of DAP1a and DAP3a structures revealed a discontinuance of the planarity in the diphenyl thioether moiety of the DAP3a most stable conformation ( Figure 2). Together with an electronic change arising from the replacement of a heteroaromatic amine with an aniline, this transformation led to the formation of molecules with very different properties. However, in in vitro assays, all compounds showed a complete loss of competition for αSYN fibrils when tested against both [ 3 H]SIL26 and [ 3 H]MODAG−001 (Table 2), invalidating our hypothesis on flexibility. Additionally, a significant decrease in affinity to Aβ was also observed, highlighting the criticality of threedimensional conformation in designing new ligands to differentiate between misfolded proteins.

Interaction with αSYN Binding Sites
Our findings demonstrated that the parent scaffolds compete with each other for the binding to αSYN as SIL26 displaces [ 3 H]MODAG−001 with a K i of 51.6 nM and MODAG−001 displaces [ 3 H]SIL26 with a K i of 10.8 nM (Figure 3). These results agreed with the previous literature [18] and aligned with the higher affinity expected for the 3,5-diphenylpyrazole, leading to the hypothesis that the two classes share the same interaction site.

Interaction with αSYN Binding Sites
Our findings demonstrated that the parent scaffolds compete with each other for the binding to αSYN as SIL26 displaces [ 3 H]MODAG−001 with a Ki of 51.6 nM and MODAG−001 displaces [ 3 H]SIL26 with a Ki of 10.8 nM (Figure 3). These results agreed with the previous literature [18] and aligned with the higher affinity expected for the 3,5diphenylpyrazole, leading to the hypothesis that the two classes share the same interaction site. The biological evaluation of the DAP hybrid compounds did not support our hypothesis. We speculate that the competition between the parent scaffolds results from the existence of two distinct binding sites partially overlapping with each other. Therefore, to successfully fit the interaction site, a different molecular hybridization approach may be necessary, e.g., the introduction of a linker connecting the two moieties or a greater overlap of the parent scaffolds to generate a smaller compound.
A recent solid-state nuclear magnetic resonance spectroscopy (ssNMR) study describes the most stable binding mode of anle138b to αSYN as the end-to-end filling of a tubular cavity [31]. This model raises concerns regarding the overall size of the DAP hybrids, which might partially obstruct their interaction with the fibrils, producing an additional cause for low binding affinity.

Chemistry
All chemicals were purchased from Sigma Aldrich (St. Louis, Missouri, USA), abcr GmbH (Karlsruhe, Germany), or Carl Roth (Karlsruhe, Germany) and used without any further purification.
Reaction progress was monitored by thin-layer chromatography (TLC) on 0.20 mm Polygram SIL G/UV254 (silica gel 60) TLC plates (Macherey-Nagel, Düren, Germany) with the chosen eluent mixture and/or analytical HPLC-MS (ESI detector, Agilent, Santa Clara, California, USA) equipped with a Luna 5 µm C18 (2)   The biological evaluation of the DAP hybrid compounds did not support our hypothesis. We speculate that the competition between the parent scaffolds results from the existence of two distinct binding sites partially overlapping with each other. Therefore, to successfully fit the interaction site, a different molecular hybridization approach may be necessary, e.g., the introduction of a linker connecting the two moieties or a greater overlap of the parent scaffolds to generate a smaller compound.
A recent solid-state nuclear magnetic resonance spectroscopy (ssNMR) study describes the most stable binding mode of anle138b to αSYN as the end-to-end filling of a tubular cavity [31]. This model raises concerns regarding the overall size of the DAP hybrids, which might partially obstruct their interaction with the fibrils, producing an additional cause for low binding affinity.

Chemistry
All chemicals were purchased from Sigma Aldrich (St. Louis, Missouri, USA), abcr GmbH (Karlsruhe, Germany), or Carl Roth (Karlsruhe, Germany) and used without any further purification.  (Compounds 14a-14c) To a solution of 13 (1.13 mmol) and the selected aryl bromide (2.27 mmol) in NMP (3.50 mL) under argon atmosphere was added Pd(PPh 3 ) 4 (5% mol). The mixture was stirred overnight at 100 • C. The crude product mixture was diluted in water and extracted with EtOAc. The organic phase was dried over MgSO 4 , evaporated under reduced pressure, and purified by flash chromatography (PE/EtOAc).  (Compounds 15a-15c) A solution of the selected 3-aryl-1-(tetrahydro-2H-pyran-2-yl)-pyrazole (0.63 mmol) in THF (2.50 mL) was cooled to −78 • C under a positive pressure of argon. A solution of n-BuLi 2.5M in hexane (0.76 mmol) was added dropwise, after which the mixture was stirred for 15 min. SnBu 3 Cl (0.76 mmol) was added dropwise, and the resulting mixture was stirred for 1 h, allowing it to reach room temperature. The reaction mixture was poured into water and extracted with EtOAc. The organic phase was dried over MgSO 4 , evaporated under reduced pressure, and purified by flash chromatography (PE/EtOAc).

General Procedure E (DAP1a-DAP3c)
To a suspension of the selected protected DAP compound (0.26 mmol) in MeOH/H 2 O 1:1 v/v (13.0 mL, 13.0 mL), HCl 37% (1.30 mL) was added, and the reaction was stirred at 80 • C for 6 h. The mixture was poured into water, neutralized with NaOH aq. 18 M, and extracted with EtOAc. The organic phase was dried over MgSO 4 , evaporated under reduced pressure, and purified by flash chromatography (PE/EtOAc or DCM/MeOH).