Symmetrical 2,7‐disubstituted 9H‐fluoren‐9‐one as a novel and promising scaffold for selective targeting of SIRT2

Abstract

Sirtuin 2 (SIRT2) belongs to the family of silent information regulators (sirtuins), which comprises nicotinamide adenine dinucleotide (NAD⁺)‐dependent protein lysine deacetylases. With a distribution across numerous tissues and organs of the human body, SIRT2 is involved in a wide range of physiological and pathological processes, such as regulating the cell cycle, energy metabolism, DNA repair, and tumorigenesis. Aberrant expression of SIRT2 has been closely associated with particular etiologies of human diseases, positioning SIRT2 as a promising therapeutic target. Herein, we detail the design overview and findings of novel symmetrical 2,7‐disubstituted 9H‐fluoren‐9‐one derivatives targeting SIRT2. SG3 displayed the most potent SIRT2‐selective inhibitory profile, with an IC₅₀ value of 1.95 $\mu \mathrm{M}$, and reduced the cell viability of human breast cancer MCF‐7 cells accompanied by hyperacetylation of α‐tubulin. Finally, molecular docking, molecular dynamics simulations, and binding free energy calculations using molecular mechanics/generalized born surface area method were performed to verify the binding ability of SG3 to SIRT2. Taken together, these results could enhance our understanding of the structural elements necessary for inhibiting SIRT2 and shed light on the mechanism of inhibition.

Lead optimization was conducted and a new scaffold of SIRT2‐selective inhibitors was discovered. The top compound SG3 selectively inhibited SIRT2 with an IC₅₀ value of 1.95 $\mu \mathrm{M}$ and reduced the MCF‐7 cell viability, accompanied by hyperacetylation of α‐tubulin. In silico studies verified the binding ability of SG3 to SIRT2.

1. INTRODUCTION

Sirtuins (SIRTs) are a phylogenetically conserved family of class III histone deacetylases (HDACs) from bacteria to humans.^{[ 1 ]} Unlike other HDACs, the catalytic activity of SIRT enzymes is mediated by the cofactor nicotinamide adenine dinucleotide (NAD⁺).^{[ 2 ]} They catalyze the removal of various acyl groups, such as acetyl, succinyl, malonyl, glutaryl, crotonyl, decanoyl, myristoyl, and palmitoyl, from histone or non‐histone substrates.^{[ 3} , ⁴ , ⁵ , ^{6 ]} Seven sirtuin isotypes (SIRT1–7) have been identified in mammals.^{[ 7 ]} SIRTs share a catalytic domain of approximately 270 amino acids, constituting a Rossmann fold, a cofactor binding domain, and a zinc‐binding domain that is not responsible for catalytic activity but is required for structural stability. They differ in their C‐ and N‐terminal domains,^{[ 8} , ^{9 ]} leading to a differentiation of their substrates, enzymatic activity, molecular targets, and cellular localization.^{[ 10} , ^{11 ]}

Unlike other SIRT isotypes, SIRT2 functions primarily in the cytoplasm but can be shifted to the nucleus during mitosis.^{[ 12 ]} SIRT2 is involved in physiological processes, such as the cell cycle, aging, energy metabolism, and genome stability, through its substrates, which include metabolic and cell cycle‐associated enzymes, cell signaling‐related substrates, transcription factors, and structural proteins.^{[ 13} , ^{14 ]} Dysregulation of SIRT2 has been implicated in a wide range of diseases, including metabolic, infectious, neurodegenerative, cardiovascular diseases, and cancer.^{[ 15 ]} Several studies have revealed that SIRT2 has a dual role in tumorigenesis, acting as a tumor promoter or suppressor.^{[ 16} , ^{17 ]} It has been demonstrated that cell type, tumor stage, and tissue‐specific effects appear to depend on the expression levels and functions of SIRT2 in the respective tissues.^{[ 18 ]} It not only affects the tumor cell cycle but can also modulate the tumor microenvironment.^{[ 19 ]} It functions as a tumor suppressor, inhibiting tumor cell growth, proliferation, angiogenesis, cell migration, and intercellular adhesion, and induces apoptosis via various factors. It also acts as a tumor promoter, increasing immune escape, invasion, energy metabolism, proliferation, and migration.^{[ 15} , ²⁰ , ²¹ , ^{22 ]} Recently, simultaneous inhibition of SIRT1, SIRT2, and SIRT3 was demonstrated to have a similar anticancer effect in several cancer cells to selective inhibition of SIRT2; however, pan‐SIRT1‐3 inhibitors caused severe toxicity issues, highlighting the advantage of SIRT2‐selective inhibitors as potential anticancer therapeutics.^{[ 23 ]} In addition, the potential of SIRT2 inhibitors to synergize with clinically approved drugs was demonstrated by investigating the role of SIRT2 inhibitors in overcoming drug resistance to dasatinib, doxorubicin, or paclitaxel, which leads to chemotherapy failure in the treatment of acute myeloid leukemia, melanoma, or certain subtypes of breast cancer cells.^{[ 24} , ²⁵ , ^{26 ]} These findings have led to considerable interest in the development of SIRT2‐selective inhibitors as novel therapeutics for the treatment of cancer,^{[ 18} , ²⁷ , ²⁸ , ²⁹ , ³⁰ , ^{31 ]} but none of the described compounds have been approved for the market due to their poor potency, selectivity, and pharmacokinetic properties,^{[ 32} , ^{33 ]} which signifies the need for novel advances in the field.

Building on our previous efforts in the design of SIRT2‐selective inhibitors, we now present the design and synthesis of a series of symmetrical 2,7‐disubstituted 9H‐fluoren‐9‐ones (SG1–SG12) as a continuation of our previous work,^{[ 34 ]} in which the structure–activity relationship (SAR) data were obtained to aid in the development of the compounds. Screening of the synthesized compounds emerged SG3 as a novel SIRT2‐selective inhibitor, and treatment of human breast cancer MCF‐7 cells with SG3 for 72 h led to a modest decrease in cell viability but a significant increase in acetylated α‐tubulin levels. Molecular docking studies against SIRT2 were performed to identify the binding modes of the compounds, followed by molecular dynamics (MD) simulations to assess the dynamic behavior and the stability of the SIRT2‐SG3 complex.

2. RESULTS AND DISCUSSION

2.1. Design

Efforts to address the development challenges focused on the design of SIRT2‐selective inhibitors have so far, resulted in almost 40 X‐ray crystal structures of SIRT2 with several different ligands.^{[ 18} , ^{35 ]} The first documented X‐ray structure of SIRT2 in complex with the analog of SirReal2, a highly selective SIRT2 inhibitor (PDB: 5DY4),^{[ 36 ]} is of great importance for the discovery of a previously unexplored region in the SIRT2 active site, the so‐called selectivity pocket, which is formed by a ligand‐induced conformational change. The dimethylpyrimidine moiety buries deeply into the selectivity pocket by forming π–π stacking contacts with Y139 and F190, while the thiazole ring occupies the extended C‐site (ECS) and the naphthyl ring protrudes into the substrate binding tunnel through π–π stacking contacts with F119, F131, and F234. The significance of H‐bonding between the carbonyl oxygen and the gatekeeper P94, mediated by the conserved water molecule (HOH549), was also emphasized (Figure 1a). Subsequently, another SIRT2 complex with an inhibitor (PDB: 5YQO)^{[ 37 ]} that features an extended binding mode inducing the formation of an enlarged hydrophobic pocket in the substrate binding channel, highlighted the additional interactions, for example, hydrophobic interactions with R97 and V233 and π–π stacking contacts with F235, contributing to the more potent inhibition of SIRT2 activity (Figure 1b).

Figure 1.

The co‐crystallized conformation of SIRT2 inhibitors from the X‐ray crystal structures of hSIRT2: (a) PDB: 5DY4, (b) PDB: 5YQO, and (c) PDB: 8OWZ. (d) The superimposed active site of three X‐ray structures of hSIRT2. H‐bonds and π–π stacking contacts are presented as yellow and cyan dashed lines, respectively.

Furthermore, to investigate the influence of the modifications at the SirReal scaffold in the lysine channel on SIRT2 affinity, new analogs were designed to extend toward the active site entrance, and the results have proposed that targeting the interaction with E237, located on the surface of SIRT2, might be a crucial strategy for the development of further potent SIRT2 inhibitors with enhanced cellular efficacy.^{[ 38 ]} The recent X‐ray structure of SIRT2 in complex with one of this type of inhibitor (PDB: 8OWZ)^{[ 35 ]} has indicated that polar interactions with R97 and H187 in the substrate channel increase and enable the additional inhibition of SIRT2. Additionally, the water‐mediated H‐bond with P94 in the cofactor binding loop has been emphasized to improve potency by stabilizing the conformation in the active site (Figure 1c).

Based on these findings, a structure‐based design strategy was employed to obtain novel SIRT2 inhibitors meeting the expectations of (i) featuring a hydrophobic moiety that optimally fills the selectivity pocket and (ii) blocking the substrate binding channel, thus the active site entrance, with more extended binding compared to our previous compounds, leading to enhanced SIRT2 efficacy.

SAR data gathered from our ongoing efforts to develop SIRT2‐selective inhibitors revealed that among the compounds obtained through previous hit‐to‐lead optimization, ST61 which features a diphenyl linked by an oxygen atom with the central phenyl ring is in a 1,4‐disubstituted arrangement, does not affect SIRT2 activity. When the terminal phenyl ring of ST61 was substituted with m‐CN to give ST62, the potency of SIRT2 inhibition was slightly increased. In the case of ST77, which was obtained by changing the 1,4‐disubstitution arrangement of ST62 to a 1,3‐disubstitution arrangement, a 2.5‐fold increase in inhibitory activity was observed.^{[ 34 ]} On the other hand, by replacing the oxygen atom linking the two phenyl rings in ST61 with a carbonyl group to yield ST60, a significant improvement in SIRT2 inhibition was achieved.^{[ 39 ]} Accordingly, the carbonyl linker and the 1,3‐disubstitution pattern of the central phenyl ring were the aspects considered to be retained in the design of new analogs due to their significant contribution to the activity. Subsequently, the fusion of the phenyl‐carbonyl linker‐phenyl motif resulted in the formation of a more rigid 9H‐fluoren‐9‐one scaffold as the central ring. Our next step was to align the binding conformation of ST77 featuring a 1,3‐disubstituted pattern with the 9H‐fluoren‐9‐one ring to pinpoint the suitable positions of this new ring system to be substituted, allowing the favorable orientation of aromatic side chains to form the desired interactions within the active site of SIRT2. According to the alignment, the second and seventh positions of the 9H‐fluoren‐9‐one ring system were identified as the optimal locations for substitution (Figure 2).

Figure 2.

The design strategy of 2,7‐disubstituted‐9H‐fluoren‐9‐ones.

Molecular docking studies were carried out to determine the aromatic side rings expected to form targeted interactions within the SIRT2 active site, as well as the linker moieties attached to positions 2 and 7 of the 9H‐fluoren‐9‐one ring. As aromatic rings, thiazole, 1‐methyl‐1H‐pyrazole, isoxazole, 1‐methyl‐1H‐imidazole, pyridine, phenyl, and 4,6‐dimethylpyrimidine were incorporated into the structure of the designed compounds, targeting the π–π interactions with the residues Y139 and F190 in the selectivity pocket, F119 at the ECS, H187 and F235 at the active site entrance. In the case of the linker fragments, urea (three‐atom) and various amide (four‐atom) moieties have been selected, offering the capability to establish H‐bonds with R97 and V233 while allowing the aromatic rings to extend toward the active site entrance and the selectivity pocket. In addition, the 9H‐fluoren‐9‐one ring was substituted with the same linker and aromatic groups at the second and seventh positions, giving the desired compounds a symmetrical character. The aim here was to achieve compounds that do not exhibit orientation diversity within the active site, driven by the desire to enhance the consistency of binding interactions, simplifying the SAR analysis and allowing for a facile synthetic strategy.

Consequently, based on the linker moiety, four types of 9H‐fluoren‐9‐ones, namely 2,7‐bis[3‐(aryl)ureido]‐9H‐fluoren‐9‐ones (SG1–SG4), 2,7‐bis[2‐oxy‐(N‐aryl)acetamido]‐9H‐fluoren‐9‐ones (SG5–SG8), 2,7‐bis[2‐(arylthio)acetamido]‐9H‐fluoren‐9‐ones (SG9–SG11), and 2,7‐bis{2‐[methyl(aryl)amino]acetamido}‐9H‐fluoren‐9‐ones (SG12) were prepared to evaluate their inhibitory potential against SIRT2.

2.2. Biological evaluation

A fluorescence‐based enzyme assay was applied to assess the ability of the target compounds (SG1–SG12) to inhibit SIRT2 activity at screening concentrations of 100 and 50 $\mu \mathrm{M}$. The inhibitory potential of the compounds against SIRT1 and SIRT3 was also determined to assess the isotype selectivity (Table 1).

Table 1.

The in vitro inhibitory activities of the compounds tested against SIRT1–3.

		Inhibition (%±SD)
		SIRT2		SIRT1		SIRT3
ID	Ar	100 $\mu \mathrm{M}$	50 $\mu \mathrm{M}$	100 $\mu \mathrm{M}$	50 $\mu \mathrm{M}$	100 $\mu \mathrm{M}$	50 $\mu \mathrm{M}$
SG1		106.21 ± 4.22	72.39 ± 2.88	22.47 ± 3.74	n.i.	44.73 ± 2.08	n.i.
SG2		48.50 ± 0.30	21.65 ± 4.19	11.31 ± 1.83	n.i.	17.50 ± 0.57	n.i.
SG3		84.39 ± 4.07	77.82 ± 1.70	n.i.	n.i.	18.61 ± 1.54	n.i.
SG4		100.75 ± 7.54	67.57 ± 3.95	24.71 ± 5.02	n.i.	42.55 ± 4.88	n.i.

		Inhibition (%±SD)
		SIRT2		SIRT1		SIRT3
ID	Ar	100 $\mu \mathrm{M}$	50 $\mu \mathrm{M}$	100 $\mu \mathrm{M}$	50 $\mu \mathrm{M}$	100 $\mu \mathrm{M}$	50 $\mu \mathrm{M}$
SG5		101.79 ± 8.24	62.39 ± 1.06	71.17 ± 5.29	53.20 ± 1.62	55.62 ± 4.42	51.66 ± 1.56
SG6		65.61 ± 3.34	64.04 ± 3.84	27.78 ± 1.31	n.i.	36.75 ± 2.26	20.51 ± 7.34
SG7		21.47 ± 2.37	n.i.	n.d.	n.d.	n.d.	n.d.
SG8		103.05 ± 2.61	99.77 ± 6.48	13.22 ± 2.22	n.i.	83.07 ± 0.27	76.37 ± 4.94

		Inhibition (%±SD)
		SIRT2		SIRT1		SIRT3
ID	Ar	100 $\mu \mathrm{M}$	50 $\mu \mathrm{M}$	100 $\mu \mathrm{M}$	50 $\mu \mathrm{M}$	100 $\mu \mathrm{M}$	50 $\mu \mathrm{M}$
SG9		71.36 ± 6.51	50.97 ± 2.28	n.i.	n.i.	90.05 ± 3.41	84.98 ± 0.10
SG10		18.13 ± 1.88	n.i.	n.d.	n.d.	n.d.	n.d.
SG11		17.24 ± 6.10	n.i.	n.d.	n.d.	n.d.	n.d.

		Inhibition (%±SD)
		SIRT2		SIRT1		SIRT3
ID	Ar	100 $\mu \mathrm{M}$	50 $\mu \mathrm{M}$	100 $\mu \mathrm{M}$	50 $\mu \mathrm{M}$	100 $\mu \mathrm{M}$	50 $\mu \mathrm{M}$
SG12		85.54 ± 5.68	74.74 ± 5.32	88.40 ± 4.45	n.i.	95.84 ± 0.52	70.09 ± 2.15

	Inhibition (%±SD)
	SIRT2		SIRT1		SIRT3
Reference compounds	100 $\mu \mathrm{M}$	50 $\mu \mathrm{M}$	100 $\mu \mathrm{M}$	50 $\mu \mathrm{M}$	100 $\mu \mathrm{M}$	50 $\mu \mathrm{M}$
Suramin	98.52 ± 0.45	78.92 ± 0.67	n.d.	n.d.	96.82 ± 3.26	26.5 ± 2.63
EX‐527	n.d.	n.d.	95.71 ± 7.75	90.23 ± 5.43	n.d.	n.d.

Abbreviations: n.d., not determined; n.i., no inhibition (inhibition % <10%); SD, Standard deviation (n = 3).

Based on the linker groups, the tested compounds can be classified into four groups for comparative evaluation of their inhibitory profile against SIRT2 activity. In general, the SIRT2 inhibitory potency of the tested compounds was enhanced, except for SG7 and SG11, supporting the assumption that the structural modifications applied in this study led to the development of 2,7‐disubstituted‐9H‐fluoren‐9‐ones, which exhibit significantly better activity compared to our previously reported lead compounds ST60 ^{[ 39 ]} and ST77.^{[ 34 ]} Among the compounds, the most potent SIRT2 inhibition was evoked by the urea‐linker type compounds (except SG2), displaying an activity range of 84%–100% and 68%–78% at 100 and 50 $\mu \mathrm{M}$, respectively, followed by the oxyacetamide‐bridged derivatives (except SG7) with an activity range of 66%–100% and 62%–100% at 100 and 50 $\mu \mathrm{M}$.

Comparable to SG1 and SG4, the oxyacetamide‐bridged counterparts SG5 and SG8 exerted similar activity, with a slight decrease observed when the three‐atom linker was extended to a four‐atom linker in compounds with a triazole ring, but a notable increase in activity by pyridine‐based derivatives under the opposite condition. In the case of isoxazole‐based analogs, SG3 exhibited effective inhibition against SIRT2, whereas the exchange of the urea group to oxyacetamide in SG7 provoked a dramatic loss of activity. However, in the presence of 1‐methyl‐1H‐pyrazole, the opposite was observed, as the urea derivative SG2 was found to be a less potent SIRT2 inhibitor than SG6, the oxyacetamide derivative.

Inspired by our previous lead compounds, 9H‐fluoren‐9‐one was linked with aryl groups, including phenyl, 1‐methyl‐1H‐imidazole, and 4,6‐dimethylpyrimidine via thioacetamide and (methylamine)acetamide moieties to yield SG9–SG11 and SG12, respectively. SG11, which bears the 4,6‐dimethylpyrimidine moiety, was the least active compound, while SIRT2 inhibition was almost negligible for the 1‐methyl‐1H‐imidazole ring‐containing derivative (SG10), consistent with previously published data.^{[ 34 ]} Indeed, the phenyl analogs SG9 and SG12 demonstrated a stronger SIRT2 inhibition, and even SG11 has a superior effect compared to SG12, confirming the more considerable impact of replacing sulfur atom with N‐methylamine in the linker moiety.^{[ 34 ]}

These findings suggested that rather than the linker or aromatic ring alone being directly linked to the activity, it is more likely that appropriate matching of the linker‐terminal aromatic ring improves the inhibitory potency of SIRT2.

The compounds were also examined for their effect on the activity of SIRT1 and SIRT3 isotypes, except for SG7, SG10, and SG11, due to their weak inhibitory potency against SIRT2. According to the results, three different types of SIRT inhibitors were identified. The compounds SG5, SG6, and SG12 were potent pan SIRT1–3 inhibitors, especially at the screening concentration of 100 $\mu \mathrm{M}$. In the case of SG8 and SG9, they were considered as the inhibitors exhibiting selectivity over SIRT2 and SIRT3 but not over SIRT1. Finally, we identified SG1, SG2, SG3, and SG4 as SIRT2‐selective inhibitors, allowing us to achieve our goal in this work.

In terms of isotype selectivity, the urea‐bridged derivatives (SG1–SG4) were the only group to exert a selectivity profile for SIRT2 that remained unsatisfactory for the other compounds. Moreover, it could be proposed that replacing a 5‐membered aromatic ring with a six‐membered one, including phenyl and pyridine in SG8, SG9, and SG12, led to enhanced inhibition of SIRT3.

As given in Table 2, the compounds identified as SIRT2‐selective inhibitors displayed a twofold greater selectivity for SIRT2 over SIRT1 as compared to SIRT2 over SIRT3. However, SG3 stands out among the compounds with excellent selectivity and potency (IC₅₀ = 1.95 $\mu \mathrm{M}$) over the enzymatic activity of SIRT2.

Table 2.

The IC₅₀ values of urea‐bridged compounds.

			Inhibition (%±SD) @100 $\mu \mathrm{M}$
ID	Ar	SIRT2 IC50 ($\mu \mathrm{M}$)	SIRT2	SIRT1	SIRT3
SG1		9.67	106.21 ± 4.22	22.47 ± 3.74	44.73 ± 2.08
SG2		n.d.	48.50 ± 0.30	11.31 ± 1.83	17.50 ± 0.57
SG3		1.95	84.39 ± 4.07	n.i.	18.61 ± 1.54
SG4		6.65	100.75 ± 7.54	24.71 ± 5.02	42.55 ± 4.88

Abbreviations: n.d., not determined; n.i., no inhibition (inhibition % <10%).

After confirming the potent and selective SIRT2 inhibition of the compounds in vitro, the most promising compounds (SG1, SG3, and SG4) were further subjected to cellular assays, and their antiproliferative activity against the MCF‐7 cell line due to having substantial SIRT2 levels.^{[ 40 ]} MCF‐7 cells were treated with the tested compounds at concentrations ranging from 1 to 200 $\mu \mathrm{M}$ for 72 h (Figure 3a). Although the compounds exerted a high level of inhibitory activity on SIRT2, the antiproliferative effect obtained was lower than expected, being almost 50% inhibition on the viability of MCF‐7 cells at the concentration of 200 $\mu \mathrm{M}$. Subsequently, western blot analysis examining SIRT2 quantification, total α‐tubulin, and acetylated α‐tubulin levels in MCF‐7 cells was performed with SG3, which exhibits the most effective antiproliferative activity, to evaluate its cell‐based activity against SIRT2 (Figure 3b). Based on the results, the in vitro SIRT2‐selective inhibitory potency of SG3 was confirmed by a notable reduction in SIRT2 presence and a 2.5‐fold increase in acetylated α‐tubulin levels in a cell‐based assay. Elevated levels of α‐tubulin acetylation have been correlated with increased microtubule stability,^{[ 41 ]} which could interfere with the dynamic instability critical for accurate mitotic spindle function in cell division. This disruption can lead to mitotic arrest and apoptosis in rapidly dividing cancer cells, making α‐tubulin hyperacetylation a valuable therapeutic target in cancer treatment.^{[ 42 ]} The ability of SG3 to induce this post‐translational modification suggests its potential to interfere with cancer cell proliferation and survival.

Figure 3.

(a) The percent cell viability in MCF‐7 cells treated with SG1, SG3, and SG4 at screening concentrations ranging from 1 to 200 $\mu \mathrm{M}$ for 72 h. Bars represent means ± SD. Each experiment was done in triplicate. Significant differences are presented as *p < 0.01, **p < 0.001, and ***p < 0.0001. (b) Representative western blot analysis in MCF‐7 cells after 72 h treatment with SG3. Significant differences are presented as ^# p < 0.05 and ^## p < 0.01.

2.3. Molecular modeling

Molecular docking studies were conducted to investigate the relationship between the inhibitory effects of compounds on SIRT2 and their interactions with the active site of SIRT2 using the Glide module implemented in the Schrödinger Small‐Molecule Drug Discovery Suite. Based on the crucial interactions emphasized within the SIRT2 structure in complex with a triazole‐based SirReal analog (PDB: 8OWZ) (Figure 1c),^{[ 35 ]} the binding patterns of SG1–SG12 were evaluated to assess whether the interactions obtained support the biological data and highlight the importance of key residues.

As regards our compounds (Figure 4, Supporting Information S2: Table S1), the accommodation of the core 9H‐fluoren‐9‐oneone ring within the ECS was mainly driven by the π–π stacking contacts or hydrophobic interactions with the residues F96 (except SG10), F119 (except SG9), F131 (except SG4 and SG9), L138, I169, F190, and I232. In particular, by examining the interactions formed by the linker groups at the second and seventh positions of this ring, it was notably observed that the urea bridge was likely to establish more tight interactions with critical residues located at the entrance of the substrate channel, for example, forming direct or water‐mediated H‐bonds with R97 and V233, which may explain why the urea‐bridged compounds (SG1–SG4) have more potent and selective inhibition on SIRT2 activity than the other derivatives. In addition, the conserved water molecule (HOH573) at the ECS allows the inhibitor to interact with P94 via water‐mediated H‐bonding, predominantly observed in thioacetamide‐bridged derivatives (SG9–SG11), followed by compounds bearing urea linkers (SG3 and SG4). In relation to the terminal aromatic rings, it was noticeable that they were all located within the selectivity pocket by forming π–π stacking contacts with Y139 and F190. The observed interactions in the selectivity pocket supported the biological data by providing a clue: SG7, which was found to be inactive among all the compounds tested, showed no interaction through its isoxazole ring, while SG10 formed π–π contacts between its 1‐methyl‐1H‐imidazol ring and only Y139. At the entrance of the active site, all types of terminal rings participated in π–π stacking or hydrophobic contacts with H187 and F235 (except SG1, SG3, and SG5), the residues targeted for interaction at this region (Supporting Information S2: Figure S1).

Figure 4.

Interaction fingerprint matrix identifying binding patterns of SG1–SG12 within the SIRT2 active site (PDB: 8OWZ).

An examination of the binding mode of SG3 (Figure 5), which was identified in biological studies as the most potent and selective SIRT2 inhibitor, revealed its position within the ECS centered on the 9H‐fluoren‐9‐one ring, and a π–π contact was observed between the ring and F190. SG3 extended beyond the selectivity pocket through its isoxazole ring, anchoring the pocket by forming π–π stacking contacts with Y139 and F190. A π–π interaction with H187 was determined for the isoxazole ring occupying the entrance of the substrate channel. The carbonyl oxygen and nitrogen atoms of the urea moiety adjacent to the substrate channel entrance engaged in H‐bonds with R97 and V233, respectively (Figure 5a). In conjunction with the aforementioned interactions, there were also hydrophobic contacts with F119, F131, L138, P140, F143, I169, I232, and F235, which contributed to the stability of the protein–ligand complex (Figure 5b). Intramolecular aromatic H‐bonds were detected between 9H‐fluoren‐9‐one and urea carbonyl (Figure 5c), similar to the intramolecular H‐bond observed for SirReal‐based SIRT2‐selective inhibitors (Figure 1a,c) between the 4,6‐dimethylpyrimidine and amide, contributing to the formation of the preferred binding conformer for enhanced potency. Although aromatic H‐bonds are generally weaker than classical H‐bonds, it has been demonstrated that they can impact molecular geometry and enhance specific conformations, thus significantly contributing to the stability of molecules, particularly in complex systems.^{[ 43} , ⁴⁴ , ^{45 ]} The binding energy of SG3 was −14.40 kcal/mol, indicating the binding strength of the SIRT2‐SG3 complex.

Figure 5.

(a) Predicted binding mode of SG3 within the active site of SIRT2 (PDB: 8OWZ). H‐bonds and π–π stacking contacts are presented by yellow and cyan dashed lines, respectively. (b) Hydrophobic contacts are represented by yellow‐green dashed lines. (c) Intramolecular aromatic H‐bonds are represented by orange dashed lines.

To assess the stability of the SIRT2‐SG3 complex, we conducted an MD simulation in line with previous studies on SIRT2.^{[ 30} , ³⁴ , ^{46 ]} The complex was placed in an orthorhombic box filled with TIP4P water molecules. The resulting solvated system was neutralized and equilibrated, followed by a 100 ns MD simulation at 300K and 1 bar pressure using the Desmond within the Schrödinger Small‐Molecule Drug Discovery Suite.

The simulation revealed that the SIRT2‐SG3 complex is stable, with root‐mean‐square deviation (RMSD) values ranging between 2 and 3 Å from the initial simulation time (Figure 6, Supporting Information S2: Figures S1 and S3). This stability is consistent with the expected behavior of the complex. As given in Supporting Information S2: Figure S2, root‐mean‐square fluctuation (RMSF) analysis showed only a single peak of mobility in the region corresponding to amino acids 240–250, a pattern observed in previous simulations.^{[ 30} , ^{34 ]} Throughout the simulation, the position of SG3 in the binding pocket gradually stabilized, optimizing its interactions within the pocket from 30 to 100 ns. Notably, the π–π stacking interactions with residues H187 and F190, identified in molecular docking studies, were maintained. The water‐mediated H‐bond with P94, as well as H‐bonds with V233, were also preserved. Additionally, hydrophobic contacts with F119, L138, I169, I232, and F235 remained intact.

Figure 6.

The RMSD values of protein (SIRT2) on the left y‐axis and ligand (SG3) on the right y‐axis over the MD simulation time. MD, molecular dynamics; RMSD, root‐mean‐square deviation.

Overall, the molecular mechanics/generalized born surface area method (MM/GBSA) analysis (Table 3) estimated the free energy of the SIRT2‐SG3 protein–ligand complex to be −86.19 kcal/mol, indicating a strong binding affinity of SG3 to the protein.

Table 3.

MM/GBSA energies for docking pose of SG3 binding at the SIRT2 active site.

ID	ΔG binding	Coulomb	Covalent	H‐bond	Lipo	Packing	Solv_GB	vdW
SG3	−86.19	−24.69	4.18	−1.25	−27.01	−4.74	36.37	−69.05

Abbreviation: MM/GBSA, molecular mechanics/generalized born surface area method.

2.4. Synthesis

The designed compounds with a 2,7‐disubstituted 9H‐fluoren‐9‐one scaffold (SG1–SG12) were synthesized according to the synthetic routes outlined in Schemes 1, 2, 3. Briefly, 2,7‐dinitro‐9H‐fluoren‐9‐one was converted to 2,7‐diamino‐9H‐fluoren‐9‐oneone (1) by SnCl₂‐mediated nitro reduction in the first step of the synthesis of 2,7‐bis[3‐(aryl)ureido]‐9H‐fluoren‐9‐ones (SG1–SG4). For the preparation of SG1–SG3, N‐aryl‐1‐carbonyl‐1H‐imidazole, the product of the substitution reaction that occurred between appropriate arylamines and 1,1′‐carbonyldiimidazole, was subjected to react with 1 to give the desired compounds. A different route was followed to synthesize SG4; 4‐nitrophenyl pyridin‐2‐ylcarbamate, the product of the reaction between 2‐aminopyridine and 4‐nitrophenyl chloroformate under basic conditions, was reacted with 1 to yield SG4 (Scheme 1).

Scheme 1.

Reagents and conditions: (i) SnCl₂·2H₂O, ethanol, reflux, 24 h; (ii) dichloromethane, rt, 1 h; (iii) 1, DMF, rt, 24 h; (iv) pyridine, dichloromethane, 0°C → rt, 24 h; (v) 1, DMF, 80°C, 24 h.

Scheme 2.

Reagents and conditions: (i) K₂CO₃, DMF, reflux, 30 min; (ii) 5% aqueous NaOH solution, reflux, 2 h; (iii) SOCl₂, reflux, 2 h; (iv) appropriate amine, TEA, dichloromethane, 0°C → rt, 6 h.

Scheme 3.

Reagents and conditions: (i) TEA, DMF, 0°C → rt, 2 h; (ii) NaOH, DMF, rt, 5 h; (iii) DIPEA, DMF, rt → 50°C, 2 h.

The synthetic pathway to the compounds featuring the structure of 2,7‐bis[2‐oxy‐(N‐aryl)acetamido]‐9H‐fluoren‐9‐one (SG5–SG8) involves nucleophilic substitution, ester hydrolysis, and condensation of a carboxylic acid derivative with an amine in the presence of thionyl chloride. First, 2,7‐dihydroxy‐9H‐fluoren‐9‐one was reacted with methyl bromoacetate to give the ester intermediate, followed by the basic hydrolysis yielding 2,2′‐[(9‐oxo‐9H‐fluorene‐2,7‐diyl)bis(oxy)]diacetic acid (2), which was then converted to the acyl chloride derivative in the presence of thionyl chloride. A condensation reaction was then carried out with the appropriate amine and acyl chloride derivative to afford the target compounds (Scheme 2).

To obtain 2,7‐bis[2‐(arylthio)acetamido]‐9H‐fluoren‐9‐ones (SG9–SG11), a two‐step process involving amidation and nucleophilic substitution reactions was employed. In the first step, the amine derivative 1 was converted to the amide derivative (3) by 2‐bromoacetyl bromide in the presence of TEA, followed by the reaction of 3 with appropriate arylthiols under basic conditions to afford the desired compounds. In the case of SG12, the nucleophilic substitution of 3 with N‐methylaniline in the presence of DIPEA was employed (Scheme 3).

Besides, the structures of the compounds were characterized by ¹H‐NMR, ¹³C‐NMR, and high‐resolution mass spectrometry (HRMS) spectra, which are presented in Supporting Information S1: Data.

3. CONCLUSION

Twelve symmetrical 2,7‐disubstituted 9H‐fluoren‐9‐one derivatives were rationally designed to identify more potent SIRT2 inhibitors. Among these, SG3 was identified as a novel selective SIRT2 inhibitor with significantly higher potency and selectivity than the lead compounds. Furthermore, a reduction in MCF‐7 cell viability is associated with an induction of α‐tubulin hyperacetylation by treatment with SG3, highlighting that its effect on cancer cell viability is mediated through SIRT2 inhibition. The binding affinity of SG3 to the SIRT2 active site was quantified using molecular docking, MD simulations, and free binding energy calculations based on the MM/GBSA approach. The results of the combined in silico studies identified structural elements required to inhibit SIRT2, which may aid in designing more successful SIRT2‐selective inhibitors through strategic molecular architecture design and optimization. Taken together, these findings may suggest that SG3 may have potential as a candidate for further development and warrants additional investigation.

4. EXPERIMENTAL

4.1. Chemistry

4.1.1. General

All chemicals were commercially purchased and used without further purification. Reactions were monitored by thin‐layer chromatography (TLC) on silica‐coated aluminum sheets (Silica gel 60 F₂₅₄, Merck) with spots visualized by UV light (254 or 366 nm). ¹H‐NMR and ¹³C‐NMR spectra (see the Supporting Information) were recorded on a Bruker Avance neo 500 MHz FT spectrometer using tetramethylsilane as the internal standard. Chemical shifts are reported in δ (ppm). All coupling constants are given in Hertz. HRMS data were acquired on a Waters LCT Premier XE mass spectrometer (high sensitivity orthogonal acceleration time‐of‐flight instrument) operating in electrospray ionization (ESI) mode, also coupled to an AQUITY Ultra Performance Liquid Chromatography system (Waters Corporation) using a UV detector monitoring at 254 nm. The purity of all target compounds was >95% according to the UPLC/MS method using (A) water + 0.1% formic acid and (B) acetonitrile + 0.1% formic acid; flow rate = 0.3 mL/min, column: Aquity BEH C18 column (2.1 × 100 mm², 1.7 mm, Waters Corporation). Melting points were determined without correction using a Stuart SMP50 automatic melting point apparatus.

The InChI codes of the investigated compounds, together with some biological activity data, are provided as Supporting Information S1.

4.1.2. Synthesis of 2,7‐diamino‐9H‐fluoren‐9‐one (1)

2,7‐Dinitro‐9H‐fluoren‐9‐one (1 mmol) and SnCl₂·2H₂O (8 mmol) were suspended in ethanol and refluxed for 24 h. At the end of the time, the reaction mixture was cooled and concentrated in vacuo. The mixture was then diluted with water, adjusted to pH 8–9 with a saturated aqueous NaHCO₃ solution, and extracted with ethyl acetate (3 × 10 mL). The combined organic layers were dried over anhydrous Na₂SO₄ and concentrated in vacuo. The crude product was used in the next step without further purification (132 mg, 87%). HRMS (ESI) calculated for C₁₃H₁₁N₂O [M+H]⁺ (m/z) 211.0871; found 211.0871.

4.1.3. Synthesis of 2,2′‐[(9‐oxo‐9H‐fluorene‐2,7‐diyl)bis(oxy)]diacetic acid (2)

The solution of 2,7‐dihydroxy‐9H‐fluoren‐9‐one (0.25 mmol), methyl bromoacetate (0.63 mmol), and K₂CO₃ (0.75 mmol) in DMF was refluxed for 30 min. At the end of the time, the cooled reaction mixture was poured into ice water, and the precipitate was filtered. The crude product, dimethyl 2,2′‐[(9‐oxo‐9H‐fluorene‐2,7‐diyl)bis(oxy)] diacetate, was refluxed in 5% aqueous NaOH solution for 2 h. After the reaction was completed, the mixture was adjusted to pH 4 with conc. HCl. The precipitate was then filtered and used in the next step without further purification (1111 mg, 74%). HRMS (ESI) calculated for C₁₇H₁₃O₇ [M+H]⁺ (m/z) 329.0661; found 329.0661.

4.1.4. Synthesis of N,N′‐(9‐oxo‐9H‐fluorene‐2,7‐diyl)bis(2‐bromoacetamide) (3)

To a solution of 2,7‐diamino‐9H‐fluoren‐9‐one (0.25 mmol) in DMF, 2‐bromoacetyl bromide (0.9 mmol) was added dropwise at 0°C and stirred at room temperature for 2 h. At the end of the time, the reaction mixture was poured into ice water, and the precipitate was filtered and used in the next step without further purification (294 mg, 64%). HRMS (ESI) calculated for C₁₇H₁₃Br₂N₂O₃ [M+H]⁺ (m/z) 450.9293; found 450.9298.

4.1.5. General synthesis method for SG1–SG3

An appropriate arylamine derivative (0.5 mmol) and 1,1′‐carbonyldiimidazole (0.6 mmol) were stirred in dichloromethane at room temperature for 1 h. At the end of the time, the reaction mixture was concentrated in vacuo. The residue was dissolved in DMF, and 1 (0.25 mmol) was added and stirred at room temperature for 24 h. Upon completion of the reaction, the mixture was poured into 0.1 M HCl solution. The precipitate was then filtered and purified by recrystallization using appropriate solvents or column chromatography on a silica gel column to produce the desired compounds.

1,1′‐(9‐Oxo‐9H‐fluorene‐2,7‐diyl)bis[3‐(thiazol‐2‐yl)urea] (SG1) was obtained from thiazol‐2‐amine according to the general method described above and purified by recrystallization from DMF. Claret red solid (51 mg, 44%). MP: >300°C. ¹H‐NMR (500 MHz, DMSO‐d ₆, 21°C, TMS): δ = 10.72 (br s, 2H, NH), 9.21 (s, 2H, NH), 7.82 (d, J = 2.0 Hz, 2H, Ar‐H), 7.62 (d, J = 8.1 Hz, 2H, Ar‐H), 7.55 (dd, J = 8.1 and 2.0 Hz, 2H, Ar‐H), 7.39 (d, J = 3.7 Hz, 2H, Ar‐H), 7.13 (d, J = 3.7 Hz, 2H, Ar‐H) ppm. ¹³C‐NMR (125 MHz, DMSO‐d ₆, 22°C, TMS): δ = 193.4 (C=O), 160.4 (Ar‐C), 152.6 (C=O), 139.8 (Ar‐C), 138.71 (Ar‐C), 137.0 (Ar‐C), 134.8 (Ar‐C), 124.8 (Ar‐C), 121.6 (Ar‐C), 114.7 (Ar‐C), and 112.9 (Ar‐C) ppm. HRMS (ESI) m/z [M+H]⁺ calcd. for C₂₁H₁₅N₆O₃S₂: 463.0647, found: 463.0648.

1,1′‐(9‐Oxo‐9H‐fluorene‐2,7‐diyl)bis[3‐(1‐methyl‐1H‐pyrazol‐3‐yl)urea] (SG2) was obtained from 1‐methyl‐1H‐pyrazol‐3‐amine according to the general method described above and purified by column chromatography on silica gel, eluting with a mixture of dichloromethane/methanol/acetic acid (95:5:0.5). Pink solid (39 mg, 34%). MP: >300°C. ¹H‐NMR (500 MHz, DMSO‐d ₆, 21°C, TMS): δ = 9.23 (br s, 2H, NH), 9.05 (s, 2H, NH), 7.82 (d, J = 1.8 Hz, 2H, Ar‐H), 7.57 (d, J = 8.2 Hz, 2H, Ar‐H), 7.55 (d, J = 2.2 Hz, 2H, Ar‐H), 7.48 (dd, J = 8.1 and 1.8 Hz, 2H, Ar‐H), 6.23 (d, J = 1.1 Hz, 2H, Ar‐H), and 3.75 (s, 6H, CH₃) ppm. ¹³C‐NMR (125 MHz, DMSO‐d ₆, 22°C, TMS): δ = 193.7 (C=O), 152.3 (C=O), 147.8 (Ar‐C), 140.5 (Ar‐C), 138.2 (Ar‐C), 134.8 (Ar‐C), 131.8 (Ar‐C), 124.3 (Ar‐C), 121.4 (Ar‐C), 114.4 (Ar‐C), 95.4 (Ar‐C), and 38.7 (CH₃) ppm. HRMS (ESI) m/z [M+H]⁺ calcd. for C₂₃H₂₁N₈O₃: 457.1737, found: 457.1739.

1,1′‐(9‐Oxo‐9H‐fluorene‐2,7‐diyl)bis[3‐(isoxazol‐3‐yl)urea] (SG3) was obtained from isoxazol‐3‐amine according to the general method described above and purified by recrystallization from DMF. Claret red solid (36 mg, 33%). MP: >300°C. ¹H‐NMR (500 MHz, DMSO‐d ₆, 25°C, TMS): δ = 9.72 (br s, 2H, NH), 9.06 (br s, 2H, NH), 8.76 (d, J = 1.8 Hz, 2H, Ar‐H), 7.81 (d, J = 2.0 Hz, 2H, Ar‐H), 7.61 (d, J = 8.0 Hz, 2H, Ar‐H), 7.50 (dd, J = 8.0 and 2.0 Hz, 2H, Ar‐H), and 6.87 (d, J = 1.8 Hz, 2H, Ar‐H) ppm. ¹³C‐NMR (125 MHz, DMSO‐d ₆, 26°C, TMS): δ = 193.4 (C=O), 160.4 (Ar‐C), 158.5 (Ar‐C), 151.8 (C=O), 139.9 (Ar‐C), 138.7 (Ar‐C), 134.8 (Ar‐C), 124.9 (Ar‐C), 121.6 (Ar‐C), 114.8 (Ar‐C), and 98.9 (Ar‐C) ppm. HRMS (ESI) m/z [M+H]⁺ calcd. for C₂₁H₁₅N₆O₅: 431.1104, found 431.1101.

4.1.6. Synthesis of 1,1′‐(9‐oxo‐9H‐fluorene‐2,7‐diyl)bis[3‐(pyridin‐2‐yl)urea] (SG4)

To a solution of 2‐aminopyridine (1 mmol) and pyridine (1 mmol) in dichloromethane, 4‐nitrophenyl chloroformate (1 mmol) was added dropwise at 0°C. The reaction mixture was then stirred at room temperature for 24 h. At the end of the time, the precipitate was filtered and washed with dichloromethane. A solution of the crude product (0.75 mmol) and 1 (0.25 mmol) in DMF was heated to 80°C for 24 h. After the completion, the reaction mixture was poured into 0.1 M HCl solution. The precipitate was then filtered and purified by recrystallization from the DMSO‐ethanol mixture. Red solid (41 mg, 36%). MP: >300°C. ¹H‐NMR (500 MHz, DMSO‐d ₆, 24°C, TMS): δ = 10.87 (br s, 2H, NH), 9.58 (s, 2H, NH), 8.32 (d, J = 4.6 Hz, 2H, Ar‐H), 7.90 (s, 2H, Ar‐H), 7.77 (t, J = 7.8 Hz, 2H, Ar‐H), 7.62–7.64 (m, 2H, Ar‐H), 7.58–7.60 (m, 2H, Ar‐H), 7.47 (d, J = 8.3 Hz, 2H, Ar‐H), and 7.02–7.05 (m, 2H, Ar‐H) ppm. ¹³C‐NMR (125 MHz, DMSO‐d ₆, 25°C, TMS): δ = 193.5 (C=O), 153.2 (C=O), 152.6 (Ar‐C), 147.4 (Ar‐C), 140.0 (Ar‐C), 139.1 (Ar‐C), 138.7 (Ar‐C), 134.9 (Ar‐C), 125.1 (Ar‐C), 121.6 (Ar‐C), 118.1 (Ar‐C), 115.0 (Ar‐C), and 112.4 (Ar‐C) ppm. HRMS (ESI) m/z [M+H]⁺ calcd. for C₂₅H₁₉N₆O₃: 451.1519, found: 451.1518.

4.1.7. General synthesis method for SG5–SG8

The solution of 2 (0.1 mmol) in SOCl₂ was refluxed for 2 h. At the end of the time, the reaction mixture was concentrated in vacuo. The acyl chloride derivative was dissolved in dichloromethane and added dropwise at 0°C to a solution of appropriate arylamine (0.2 mmol) and triethylamine (0.3 mmol) in dichloromethane. The reaction mixture was then stirred at room temperature for 6 h. After the completion of the reaction, the mixture was concentrated in vacuo. The residue was washed with 1% aqueous NaHCO₃ solution and filtered, followed by purification by recrystallization using appropriate solvents.

2,2′‐[(9‐Oxo‐9H‐fluorene‐2,7‐diyl)bis(oxy)]bis[N‐(thiazol‐2‐yl)acetamide] (SG5) was obtained from thiazol‐2‐amine according to the general method described above. The precipitate was washed with DMF and purified by recrystallization from the DMSO‐acetone mixture. Pink solid (42 mg, 21%). MP: >300°C. ¹H‐NMR (500 MHz, DMSO‐d ₆, 23°C, TMS): δ = 12.39 (s, 2H, NH), 7.61 (d, J = 8.0 Hz, 2H, Ar‐H), 7.51 (d, J = 3.6 Hz, 2H, Ar‐H), 7.27 (d, J = 3.6 Hz, 2H, Ar‐H), 7.15–7.18 (m, 4H, Ar‐H), and 4.96 (s, 4H, CH₂) ppm. ¹³C‐NMR (125 MHz, DMSO‐d ₆, 23°C, TMS): δ = 193.1 (C=O), 167.0 (C=O), 162.8 (Ar‐C), 158.8 (Ar‐C), 157.8 (Ar‐C), 137.8 (Ar‐C), 135.5 (Ar‐C), 122.1 (Ar‐C), 121.6 (Ar‐C), 114.4 (Ar‐C), 110.8 (Ar‐C), and 66.8 (CH₂) ppm. HRMS (ESI) m/z [M+H]⁺ calcd. for C₂₃H₁₇N₄O₅S₂: 493.0640, found: 493.0658.

2,2′‐[(9‐Oxo‐9H‐fluorene‐2,7‐diyl)bis(oxy)]bis[N‐(1‐methyl‐1H‐pyrazol‐3‐yl) acetamide] (SG6) was obtained from 1‐methyl‐1H‐pyrazol‐3‐amine according to the general method described above and purified by recrystallization from the DMSO‐methanol mixture. Orange solid (63 mg, 25%). MP: 276.1–276.5°C. ¹H‐NMR (500 MHz, DMSO‐d ₆, 22°C, TMS): δ = 10.59 (m, 2H, NH), 7.57–7.60 (m, 4H, Ar‐H), 7.13–7.15 (m, 4H, Ar‐H), 6.64 (s, 2H, Ar‐H), 4.76 (s, 4H, CH₂), and 3.75 (s, 6H, CH₃), ppm. ¹³C‐NMR (125 MHz, DMSO‐d ₆, 23°C, TMS): δ = 193.1 (C=O), 165.8 (C=O), 158.9 (Ar‐C), 146.5 (Ar‐C), 137.7 (Ar‐C), 135.5 (Ar‐C), 131.6 (Ar‐C), 122.0 (Ar‐C), 121.6 (Ar‐C), 110.8 (Ar‐C), 97.2 (Ar‐C), 67.3 (CH₂), and 38.8 (CH₃) ppm. HRMS (ESI) m/z [M+H]⁺ calcd. for C₂₅H₂₃N₆O₅: 487.1730, found: 487.1749.

2,2′‐[(9‐Oxo‐9H‐fluorene‐2,7‐diyl)bis(oxy)]bis{N‐[3‐(isoxazol‐3‐yl)]acetamide} (SG7) was obtained from isoxazol‐3‐amine according to the general method described above. The precipitate was washed with DMF and purified by recrystallization from the DMSO‐acetone mixture. Pink solid (37 mg, 20%). MP: >300°C. ¹H‐NMR (500 MHz, DMSO‐d ₆, 22°C, TMS): δ = 11.30 (s, 2H, NH), 8.83 (d, J = 2.0 Hz, 2H, Ar‐H), 7.61 (d, J = 8.2 Hz, 2H, Ar‐H), 7.14–7.17 (m, 4H, Ar‐H), 6.93 (s, 2H, Ar‐H), and 4.87 (s, 4H, CH₂) ppm. ¹³C‐NMR (125 MHz, DMSO‐d ₆, 23°C, TMS): δ = 193.1 (C=O), 167.3 (C=O), 160.9 (Ar‐C), 158.8 (Ar‐C), 157.5 (Ar‐C), 137.7 (Ar‐C), 135.5 (Ar‐C), 122.1 (Ar‐C), 121.6 (Ar‐C), 110.9 (Ar‐C), 99.7 (Ar‐C), and 67.2 (CH₂) ppm. HRMS (ESI) m/z [M+H]⁺ calcd. for C₂₃H₁₇N₄O₇: 461.1097, found: 461.1086.

2,2′‐[(9‐Oxo‐9H‐fluorene‐2,7‐diyl)bis(oxy)]bis[N‐(pyridin‐2‐yl)acetamide] (SG8) was obtained from 2‐aminopyridine according to the general method described above and purified by recrystallization from the DMSO‐acetone mixture. Orange solid (45 mg, 23%). MP: >300°C. ¹H‐NMR (500 MHz, DMSO‐d ₆, 26°C, TMS): δ = 10.57 (s, 2H, NH), 8,35 (d, J = 4.9 Hz, 2H, Ar‐H), 8.05 (d, J = 8.3 Hz, 2H, Ar‐H), 7.81 (dt, J = 7.9 and 1.9 Hz, 2H, Ar‐H), 7.59–7.61 (m, 2H, Ar‐H), 7.13–7.16 (m, 6H, Ar‐H), and 4.89 (s, 4H, CH₂) ppm. ¹³C‐NMR (125 MHz, DMSO‐d ₆, 24°C, TMS): δ = 193.1 (C=O), 167.5 (C=O), 158.9 (Ar‐C), 151.8 (Ar‐C), 148.6 (Ar‐C), 138.9 (Ar‐C), 137.7 (Ar‐C), 135.6 (Ar‐C), 122.1, (Ar‐C), 121.7 (Ar‐C), 120.3 (Ar‐C), 114.2 (Ar‐C), 110.7 (Ar‐C), and 67.4 (CH₂) ppm. HRMS (ESI) m/z [M+H]⁺ calcd. for C₂₇H₂₁N₄O₅: 481.1512, found: 481.1504.

4.1.8. General synthesis method for SG9–SG11

An appropriate arylthiol (0.4 mmol) and NaOH (0.4 mmol) were stirred in DMF at room temperature for 15 min. At the end of the time, 3 (0.1 mmol) was added and stirred at room temperature for 5 h. Upon completion of the reaction, the mixture was poured into ice water. The precipitate was then filtered and purified by recrystallization from the DMSO‐ethanol mixture.

N,N′‐(9‐Oxo‐9H‐fluorene‐2,7‐diyl)bis[2‐(phenylthio)acetamide] (SG9) was obtained from thiophenol according to the general method described above. Pink solid (47 mg, 46%). MP: 243.4–243.9°C. ¹H‐NMR (500 MHz, DMSO‐d ₆, 23°C, TMS): δ = 10.45 (s, 2H, NH), 7.87 (d, J = 1.4 Hz, 2H, Ar‐H), 7.61–7.66 (m, 4H, Ar‐H), 7.42 (d, J = 7.4 Hz, 4H, Ar‐H), 7.34 (t, J = 7.7 Hz, 4H, Ar‐H), 7.22 (t, J = 7.4 Hz, 2H, Ar‐H), and 3.88 (s, 4H, CH₂) ppm. ¹³C‐NMR (125 MHz, DMSO‐d ₆, 24°C, TMS): δ = 193.1 (C=O), 167.6 (C=O), 139.9 (Ar‐C), 139.4 (Ar‐C), 136.1 (Ar‐C), 134.7 (Ar‐C), 129.5 (Ar‐C), 128.7 (Ar‐C), 126.6 (Ar‐C), 125.4 (Ar‐C), 121.7 (Ar‐C), 115.3 (Ar‐C), and 38.0 (CH₂) ppm. HRMS (ESI) m/z [M+H]⁺ calcd. for C₂₉H₂₃N₂O₃S₂: 511.1150, found: 511.1152.

N,N′‐(9‐Oxo‐9H‐fluorene‐2,7‐diyl)bis{2‐[(1‐methyl‐1H‐imidazol‐2‐yl)thio]acetamide} (SG10) was obtained from 1‐methyl‐1H‐imidazole‐2‐thiol according to the general method described above. Pink solid (85 mg, 55%). MP: >300°C. ¹H‐NMR (500 MHz, DMSO‐d ₆, 29°C, TMS): δ = 10.59 (s, 2H, NH), 7.86 (s, 2H, Ar‐H), 7.62 (s, 4H, Ar‐H), 7.26 (s, 2H, Ar‐H), 6.98 (s, 2H, Ar‐H), 3.89 (s, 4H, CH₂), and 3.62 (s, 6H, CH₃) ppm. ¹³C‐NMR (125 MHz, DMSO‐d ₆, 23°C, TMS): δ = 193.2 (C=O), 167.3 (C=O), 139.9 (Ar‐C), 139.4 (Ar‐C), 134.7 (Ar‐C), 129.0 (Ar‐C), 125.3 (Ar‐C), 124.1 (Ar‐C), 121.7 (Ar‐C), 115.2 (Ar‐C), 38.8 (CH₂), and 33.5 (CH₃) ppm. HRMS (ESI) m/z [M+H]⁺ calcd. for C₂₅H₂₃N₆O₃S₂: 519.1273, found: 519.1276.

N,N′‐(9‐Oxo‐9H‐fluorene‐2,7‐diyl)bis{2‐[(4,6‐dimethylpyrimidin‐2‐yl)thio]acetamide} (SG11) was obtained from 4,6‐dimethylpyrimidin‐2‐thiol according to the general method described above. Pink solid (34 mg, 35%). MP: 290.1–290.3°C. ¹H‐NMR (500 MHz, DMSO‐d ₆, 37°C, TMS): δ = 10.43 (s, 2H, NH), 7.87 (s, 2H, Ar‐H), 7.69 (d, J = 7.5 Hz, 2H, Ar‐H), 7.62 (d, J = 8.1 Hz, 2H, Ar‐H), 6.96 (s, 2H, Ar‐H), 4.06 (s, 4H, CH₂), and 2.34 (s, 12H, CH₃) ppm. ¹³C‐NMR (125 MHz, DMSO‐d ₆, 24°C, TMS): δ = 193.3 (C=O), 169.7 (C=O), 167.4 (Ar‐C), 167.4 (Ar‐C), 140.1 (Ar‐C), 139.3 (Ar‐C), 134.7 (Ar‐C), 125.3 (Ar‐C), 121.6 (Ar‐C), 116.6 (Ar‐C), 115.2 (Ar‐C), 36.0 (CH₂), and 23.8 (CH₃) ppm. HRMS (ESI) m/z [M+H]⁺ calcd. for C₂₉H₂₇N₆O₃S₂: 571.1586, found: 571.1581.

4.1.9. Synthesis of N,N′‐(9‐oxo‐9H‐fluorene‐2,7‐diyl)bis{2‐[methyl(phenyl)amino]acetamide} (SG12)

N‐Methylaniline (0.6 mmol) and diisopropylethylamine (0.75 mmol) were stirred in DMF at room temperature for 15 min. At the end of the time, 3 (0.25 mmol) was added and stirred while heating the mixture to 50°C. After 2 h, the cooled reaction mixture was poured into ice water. The precipitate was then filtered and purified by recrystallization from the DMSO‐ethanol mixture. Reddish brown solid (50 mg, 40%). MP: 223.7–224.8°C. ¹H‐NMR (500 MHz, DMSO‐d ₆, 36°C, TMS): δ = 10.16 (s, 2H, NH), 7.89 (d, J = 1.7 Hz, 2H, Ar‐H), 7.74 (dd, J = 8.1 and 1.9 Hz, 2H, Ar‐H), 7.61 (d, J = 8.1 Hz, 2H, Ar‐H), 7.18 (t, J = 7.4 Hz, 4H, Ar‐H), 6.71 (d, J = 8.1 Hz, 2H, Ar‐H), 6.66 (t, J = 7.2 Hz, 4H, Ar‐H), 4.16 (s, 4H, CH₂), and 3.06 (s, 6H, CH₃) ppm. ¹³C‐NMR (125 MHz, DMSO‐d ₆, 24°C, TMS): δ = 193.3 (C=O), 169.7 (C=O), 149.7 (Ar‐C), 139.9 (Ar‐C), 139.2 (Ar‐C), 134.7 (Ar‐C), 129.4 (Ar‐C), 125.5 (Ar‐C), 121.5 (Ar‐C), 116.8 (Ar‐C), 115.5 (Ar‐C), 112.4 (CH₂), and 56.3 (CH₃) ppm. HRMS (ESI) m/z [M+H]⁺ calcd. for C₃₁H₂₉N₄O₃: 505.2240, found: 505.2242.

4.2. Biological assays

4.2.1. Enzyme inhibition assay

The SIRT inhibition assays were carried out using SIRT Direct Fluorescent Screening Assay Kits (Item no: 10010401 for SIRT1; ab133082 for SIRT2; 10011566 for SIRT3), according to the protocol provided by the manufacturers (Cayman Chemical and Abcam). Briefly, test compound solution in DMSO (5 µL/well), SIRT assay buffer (25 µL/well), human recombinant SIRT1/2/3 enzymes (5 µL/well), and substrate solution containing direct NAD⁺ and direct peptide (15 µL/well) were incubated on a shaker for 45 min at 37°C. After adding developer solution (50 µL/well), including direct developer and direct nicotinamide, to stop the reaction, the incubation was continued for 30 min at room temperature. Fluorescence readings were obtained using a SpectraMax i3x Multi‐Mode Microplate Reader with an excitation wavelength of 360 nm and an emission of 460 nm. The value of percent inhibition was calculated from the fluorescence readings of inhibited wells relative to those of control wells. Each experiment was carried out in triplicates, and the concentrations of the compounds for IC₅₀ determined were in the range from 0.01 to 100 $\mu \mathrm{M}$. The IC₅₀ values were calculated using logarithmic non‐linear regression curves in GraphPad Prism 9.0 (GraphPad Software Inc.).

4.2.2. Antiproliferative activity assay

The human breast cancer MCF‐7 cells were purchased from ATCC, and the cells were cultured in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% FBS and 1% penicillin‐streptomycin until 70% confluency was reached. The cytotoxic effects of synthesized compounds on MCF‐7 cells were evaluated by MTT assay.^{[ 47} , ^{48 ]} A 180 µL of 5 × 10⁴ cell/mL cell suspension was seeded on a 96‐well plate and incubated for 24 h at 37°C in a 5% CO₂ humidified atmosphere. The cells were then treated with the compounds at different concentrations (1, 10, 25, 50, 100, and 200 $\mu \mathrm{M}$) and incubated for 72 h. Following the incubation, the cells were treated with MTT reagent (5 mg/mL) for 4 h, and the absorbance of reduced formazan crystals was measured by a spectrophotometer (Thermo) at 540 nm. The untreated cells were used as a control. Each experiment was performed in triplicate, and the IC₅₀ values were calculated using linear regression equations obtained from the concentration versus percent viable cell amount graphs. Statistical analyses were performed using GraphPad Prism (GraphPad Software Inc.). Data obtained from the cell culture experiments were expressed as mean ± SD, and a one‐way ANOVA test was applied for multiple comparisons.

4.2.3. Western blot analysis

Cells were planted at 2.5 × 10⁶ cells/mL and treated with the compounds for 72 h. After the incubation, cells were lysed with a cell lysis buffer (Cell Signaling); 25 µg of protein samples was applied on a 10% SDS‐polyacrylamide gel and transferred to a nitrocellulose membrane, and 5% nonfat dry milk in PBS buffer was used to block nonspecific binding sites. Membranes were incubated overnight with primary antibodies (Cell Signaling) at 4°C and then at room temperature with IRDye secondary antibodies for 1 h. Signals were detected with the ChemiDoc Imaging System (Bio‐Rad). The band intensities were quantified using Image Lab software (Bio‐Rad). β‐Actin was used for normalization, and the data were expressed as protein expression levels.

4.3. Molecular docking

Molecular docking studies were carried out using the Induced Fit Docking (IFD) protocol^{[ 49 ]} implemented in the Schrödinger Small‐Molecule Drug Discovery Suite (Small‐Molecule Drug Discovery Suite, 2024‐2, Schrödinger, LLC, New York, NY, 2024). The IFD protocol includes docking the compounds with Glide docking, Prime Refinement,^{[ 50 ]} and calculating the binding score by Glide redocking of the protein–ligand complex. Compounds that were built through the builder panel in Maestro were subjected to ligand preparation in LigPrep (Schrödinger Release, 2024‐2: LigPrep, Schrödinger, LLC, New York, NY, 2024) using default conditions. X‐ray crystal structure of hSIRT2 in complex with a triazole‐based SirReal analog (PDB: 8OWZ)^{[ 35 ]} was retrieved from the RCSB Protein Data Bank (rcsb.org). The protein was prepared using the Protein Preparation Wizard tool. Water molecules were deleted except for HOH573 and HOH676, hydrogen atoms were added, and all‐atom charges and atom types were assigned, followed by the energy minimization and refinement of the protein to 0.20 Å RMSD by applying the OPLS4 force field. The centroid of the X‐ray ligand was defined as a box of similar size to the co‐ligand. Glide docking was performed employing standard precision (SP) mode, using OPLS4 force field and default Van der Waals (vdW) radius scaling factors of 0.50. Twenty poses for ligands were saved for each docking run. Prime Refinement is performed on residues within 5 Å of the ligand pose, whereas Glide redocking was performed using SP docking mode and the OPLS4 all‐atom force field. The best pose for each compound was selected considering the lowest Glide GScore and the formation of the essential interactions.

4.4. Molecular dynamics

The MD protocol utilized in prior studies^{[ 30} , ³⁴ , ^{46 ]} was used to analyze the docked SIRT2‐SG3 complex. The protocol employed was executed using the Desmond module within the Schrödinger Small‐Molecule Drug Discovery Suite (Small‐Molecule Drug Discovery Suite 2018‐1, Schrödinger, LLC, New York, NY, 2018). The simulation system was prepared to mimic physiological conditions, employing explicit solvent modeling with the TIP4P water model and an orthorhombic box measuring 10 × 10 × 10 ų, supplemented with 0.15 M salt. To ensure the electrical neutrality of the entire simulation system, counter ions, specifically Na⁺ and Cl⁻, were added at a minimum distance of 20 Å from the ligand within the simulation system. Subsequently, the system underwent relaxation in multiple stages: (i) initial relaxation involved up to 2000 minimization steps with a force constant of 50 kcal/mol Ų, applying harmonic restraints to the solute atoms; (ii) a 12 ps MD simulation was conducted at 10K with a force constant of 50 kcal/mol Ų. This was carried out under the NPT ensemble, Berendsen thermostat, and barostat, all while retaining harmonic restraints; (iii) the system was then heated from 10 to 300K over a 24 ps period, with harmonic restraints still in place. The NPT ensemble, Berendsen thermostat, and barostat were used for this phase; (iv) a 24 ps MD simulation at 300K without harmonic restraints was performed using the NPT ensemble, Nose‐Hoover thermostat, and Martyna‐Tobias‐Klein barostat. Following the system's relaxation, the docked complex underwent a 100 ns MD simulation using default parameters under the NPT ensemble. Upon completion of the MD simulations, the resulting trajectories were meticulously analyzed, and key parameters, including RMSD, RMSF, and protein–ligand contacts, were computed using simulation interaction diagrams.

4.5. MM/GBSA study

The ligand‐residue free energies of binding calculations were performed using the MM/GBSA method in the Schrödinger Suite. The following equation is used to calculate the binding energy:

$$\Delta G_{\text{bind}}=\Delta E_{\text{MM}}+\Delta G_{\text{solv}}+\Delta G_{\text{SA}},$$

where ΔE _MM is the difference in minimized energies:

$$\Delta E_{\text{MM}}=E_{\left(\text{complex}\right)}\hspace{0.17em}–\hspace{0.17em}{E}_{\left(\text{ligand}\right)}\hspace{0.17em}–\hspace{0.17em}{E}_{\left(\text{receptor}\right)}.$$

The difference in GBSA solvation energy of the complex and the sum of ligand and protein solvation energies is denoted by ΔG _solv. Also, ΔG _SA is the difference in surface area energy of the complex and the sum of protein and ligand. A script was used to calculate the average MM/GBSA binding energy, which also generates Coulomb energy (Coulomb), covalent binding energy (Covalent), hydrogen‐bonding energy (H‐bond), lipophilic energy (Lipo), Generalized Born electrostatic solvation energy (Solv_GB), and van der Waals energy (vdW).

CONFLICTS OF INTEREST STATEMENT

The authors declare no conflicts of interest.

Supporting information

Spectral data (¹H‐NMR, ¹³C‐NMR, and HRMS spectra) of the compounds entitled SG1‐SG12.

Molecular docking (Table S1. Predicted binding conformation of the docked compounds (SG1, SG2, SG4‐SG12) within the SIRT2 active site), Molecular dynamics simulations (Figure S1. The RMSD for Cα atoms (Å) with respect to the initial structures as functions of simulation time (ns) for the SIRT2 protein.; Figure S2. The RMSF for Cα atoms (Å) concerning the initial structures as functions of simulation time (ns) for the SIRT2 protein.; Figure S3. The RMSD for heavy atoms (Å) with respect to the initial structures as functions of simulation time (ns) for the ligands.; Figure S4. A timeline representation of the interactions and protein‐ligand contacts).

Kaya S. G., Eren G., Massarotti A., Gunindi H. B., Bakar‐Ates F., Ozkan E., Arch. Pharm. 2024;357:e2400661. 10.1002/ardp.202400661

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.