In Vitro Anticancer Activity and In Silico Target Profiling of 5-(Piperazin-1-ylsulfonyl)-1,3-oxazole-4-carbonitriles

TL;DR

This study identifies new oxazole-based compounds with strong anticancer activity and good selectivity, particularly in neuroblastoma cells.

Contribution

The discovery of 5-(piperazin-1-ylsulfonyl)-1,3-oxazole-4-carbonitriles as selective anticancer agents with potential kinase-targeting mechanisms.

Findings

Compound 7b showed high activity in neuroblastoma cells with low toxicity to non-cancerous cells.

Molecular docking predicted strong interactions with Aurora A kinase and cyclin-dependent kinases.

ADMET properties of lead compounds were favorable compared to doxorubicin, though biodegradability was limited.

Abstract

Sulfonylated 5-piperazine-substituted 1,3-oxazole-4-carbonitriles were synthesized and evaluated for in vitro anticancer activity. Cytotoxicity was assessed in hepatocellular (HepG2, Huh7), breast (MCF-7, MDA-MB-231), cervical (HeLa), melanoma (M21), and neuroblastoma (Kelly, SH-SY5Y) cell lines, with HEK293 cells used as a non-malignant control. Compounds 7a, 7b, and 8aa emerged as lead structures. Notably, compound 7b showed the highest activity in Kelly neuroblastoma cells (IC50 = 1.3 µM) while exhibiting low cytotoxicity toward HEK293 cells (IC50 > 10 µM), indicating an improved selectivity profile relative to doxorubicin. In silico molecular docking suggested favorable interactions of the lead compounds with several cancer-associated proteins, with the highest predicted affinity observed for Aurora A kinase, along with additional predicted interactions with cyclin-dependent kinases. Predicted ADMET properties of compounds 7a, 7b, and 8aa compared favorably with doxorubicin, although the lead compounds were not readily biodegradable under OECD 301D conditions. Overall, these findings identify oxazole-4-carbonitriles as promising anticancer candidates with a putative kinase-directed mechanism of action.

Figures

Full text

1. Introduction

First reported in 1917, oxazoles became important after the discovery of penicillin antibiotics [1]. Structurally, oxazoles are planar, sp^2^-hybridized five-membered rings with oxygen at position 1 and nitrogen at position 3, separated by carbon; see Figure 1 center. Their structure is isomeric to furan and pyridine. The electronic properties of oxazoles affect their reactivity, including participation in electrophilic and Diels–Alder reactions. Compared to pyridine, oxazoles are less resistant to oxidation, but more stable in acid and at elevated temperatures [2], facilitating their use in organic synthesis and drug development [3]. Their aromatic structure allows for substitutions at C-2, C-4, and C-5 to tune electronic, steric, and lipophilic properties. Oxazoles act as bioisosteres of amide or peptide bonds, enabling hydrogen bonding and molecular recognition, and serve as drug-development building blocks via oxygen- and nitrogen-mediated interactions with enzymes and receptors [3].

1,3-Oxazoles inhibit enzymes, block receptor activity, or interfere with microbial biosynthesis, supporting use as anticancer, antiviral, anti-inflammatory, and antibacterial agents; see Figure 2 [4]. Oxazole-containing drugs have been shown to demonstrate anticancer activity against multiple cancer types. Tivozanib, for example, binds the ATP-binding site and regulatory domain of VEGFR2, blocking autophosphorylation and signal transduction [5]. Also, the sulfonyl and cyano groups [5], and the piperazine moiety are present in various clinically used anticancer drugs: Prexasertib (Chk1 and Chk2 inhibitor) [6], Olaparib (PARP inhibitor) [6], Palbociclib (selective inhibitor of the cyclin-dependent kinases CDK4 and CDK6) [7], and Navitoclax (Bcl-2 inhibitor) [8] are the examples.

Cancer is the second leading cause of death worldwide. In 2018, it was responsible for nearly 9.6 million deaths and about 18.1 million new cases [9]. According to the 2025 report, cancer causes about 2.04 million new cases and 618 000 deaths each year in the United States, reflecting a rise among younger adults and women [10]. Kinase signaling is central in sustaining oncogenic development [11]. Thus, oxazoles are used in kinase inhibitor design, being potent inhibitors of class III receptor tyrosine kinases such as FLT3 (wild-type and mutant), FLT4, and c-KIT, with demonstrated activity in non-small-cell lung cancer [12]. Cyclin-dependent kinases (CDKs) regulate cell-cycle progression and transcription; however, high conservation of the ATP-binding site limits inhibitor selectivity [13]. The oxazole-based CDK7/9 inhibitor SNS-032 blocks RNA polymerase II phosphorylation, inducing transcriptional shutdown and cell death [14,15]. Oxazoles also target histone deacetylases 2 and 6) [15], poly(ADP-ribose) polymerase-2 (PARP-2) [16], and matrix metalloproteinases MMP-2 and MMP-9 [17], with antiproliferative effects in cervical, hepatocellular, breast, melanoma, and neuroblastoma cancers.

Cancer growth is associated with oncogene overexpression [18] and tumor suppressor inactivation [19], with MYCN amplification occurring in 20–25% of cases and correlating with aggressive disease [20]. N-MYC, one of three MYC subtypes proteins (C-MYC, L-MYC, and N-MYC), encodes a transcription factor [21] that regulates cell cycle, growth, apoptosis, DNA repair, metabolism, and immune response [22]. The stability of MYC protein is controlled by phosphorylation at the conserved MBI sequence, which directs proteolysis and ubiquitination [23]. Dephosphorylation of N-MYC at Ser62 by protein phosphatase 2A allows the SCF-FbxW7 E3 ubiquitin ligase to add a K48-linked ubiquitin chain [24]. Aurora-A kinase blocks this process, preventing metabolic degradation via an inhibitor-sensitive N-MYC/Aurora-A complex, leading to MYCN accumulation [25].

CD532 (IC_50_ = 45 nM; Figure 3, left) both inhibits Aurora A kinase activity and promotes MYCN degradation [26]. Alisertib (MLN8237; Figure 3, center) is an ATP-competitive Aurora A kinase inhibitor (IC_50_ = 1.2 nM) that disrupts cell cycle progression, induces apoptosis, and suppresses MYCN expression [27]. CCT137690 (Figure 3, right) is an isoxazole-containing compound that inhibits Aurora A (IC_50_ = 15 nM), Aurora B (IC_50_ = 25 nM), and Aurora C (IC_50_ = 19 nM), resulting in tumor growth inhibition in MYCN-overexpressing neuroblastoma models. These examples illustrate the therapeutic relevance of Aurora A kinase in neuroblastoma, while also highlighting the need for structurally diverse scaffolds capable of modulating Aurora A-associated pathways. In this study, a series of oxazole–piperazine–sulfonyl hybrids were synthesized and evaluated for their in vitro anticancer activity, with a primary focus on neuroblastoma models. Initial cytotoxicity screening identified compounds exhibiting selective activity against MYCN-amplified Kelly and SH-SY5Y neuroblastoma cell lines while sparing non-malignant HEK293 cells. The most active compounds (7a, 7b, and 8aa) were further characterized by dose–response analyses and selectivity assessment, followed by evaluation of the lead compound 7b across a broader panel of human cancer cell lines, including HeLa, HepG2, Huh7, MDA-MB-231, MCF7, and M21. To rationalize the observed biological profile, potential molecular targets were explored using in silico docking studies, and the pharmacokinetic properties of the lead compounds were assessed by ADME analysis.

2. Result and Discussion

2.1. Chemistry

Previously, we synthesized and studied sulfonyl derivatives of oxazole, which demonstrated significant antitumor activity. In our study several 5-arylsulfonyl-1,3-oxazole-4-carbonitriles exhibited high activity on the CNS cancer cell line [26]. In particular, the compound 2-(4-fluorophenyl)-5-(toluene-4-sulfonyl)-1,3-oxazole-4-carbonitrile demonstrated growth inhibition against glioblastoma cells SF-268 (GI_50_ = 4.61µM) and SF-539 (GI_50_ = 2.44 µM) and indicate low cytotoxicity with values LC_50_ >100 µM. Therefore, the anticancer activity of sulfonyl derivatives of 2-phenyl-4-cyano-1,3-oxazoles can be improved by inserting a piperazine moiety having multiple biotargets.

Synthesis of the target 5-(piperazin-1-yl)oxazole-4-carbonitriles 2a–2c and 3aa–3ce was carried out as follows (Scheme 1). First, the appropriate 2-acylamino-3,3-dichloroacrylonitriles I were dissolved in anhydrous THF, in the presence of triethylamine, and then Boc-protected piperazine was added, and the reaction mixture was stirred at 20–25 C for 48h. To remove the Boc-protecting group, the intermediate products IA were dissolved in dioxane, and hydrogen chloride gas was bubbled through the solution for 30 min. The reaction mixture was then left to stand for 24h. With the corresponding 5-(piperazin-1-yl)oxazole-4-carbonitrile hydrochlorides (2a–2c) in solution 1,4-dioxane, triethylamine was added, followed by the appropriate sulfonyl chloride. The reaction mixture was stirred and refluxed for 8h. The final products 3aa–3ce were purified by recrystallization from EtOH.

The structure-synthesized compound 2a–2c and 3aa–3ce is shown in Figure 4.

The target 5-(piperazin-1-ylsulfonyl)oxazole-4-carbonitriles were synthesized according to the procedure (Scheme 2). The corresponding 2-acylamino-3,3-dichloroacrylonitriles I were used as starting materials. Their reaction with an excess of NaSH in methanol solution led to cyclization, yielding substituted 5-mercaptooxazoles II. Alkylation of compounds 2 with benzyl chloride in ethanol in the presence of triethylamine afforded substituted 1,3-oxazoles 3 bearing a benzylthio group at position 5 of the oxazole ring. Oxidative chlorination of compounds III in aqueous acetic acid at 0–5 °C produced 2-aryl-4-cyano-1,3-oxazole-5-sulfonyl chlorides IV. Reaction of these intermediates with 1-phenylpiperazine in dioxane in the presence of excess triethylamine gave the corresponding 2-aryl-5-(4-phenylpiperazin-1-yl)sulfonyl-1,3-oxazole-4-carbonitriles 5a, 5b (Scheme 2).

Next, the substituted 4-cyano-1,3-oxazole-5-sulfonyl chlorides 4 with Boc-piperazine gave the corresponding Boc-protected sulfonamides 6. Passing gaseous HCl through a dioxane solution of sulfonamides 6 resulted in the removal of the Boc protecting group and the formation of sulfonamide hydrochlorides 7a, 7b. Finally, the reaction of these compounds with the corresponding sulfonyl chlorides yielded sulfonamides 8aa–8bd, containing two sulfonyl groups connected by a piperazine linker (Scheme 3) and structure-synthesized compound 7a–7b and 8aa–8bd shown in Figure 5.

The structure of the compounds 2a–2c, 3aa–3ce, 5a–5b, 7a–7b and 8aa–8bd was confirmed using the ^1^H, ^13^C NMR and FTIR spectroscopy, and elemental analysis (see Supporting Information Figures S1–S96b).

2.2. Biology (In Vitro)

To evaluate the anticancer potential of the synthesized compounds, preliminary cytotoxicity was conducted at a fixed concentration of 2.5 µM using the WST-1 assay in a neuroblastoma Kelly cells and non-malignant human embryonic kidney cell line HEK293, shown in Figure 6. An initial cytotoxicity screening was designed to assess both anticancer activity and selectivity of the synthesized compounds. The screening was performed at a fixed concentration of 2.5 µM using the WST-1 assay in neuroblastoma Kelly cells, representing a malignant model, and non-malignant human embryonic kidney HEK293 cells as a control (Figure 6).

Neuroblastoma Kelly cells exhibited pronounced sensitivity to a subset of the synthesized compounds, with compounds 7a, 7b, and 8aa inducing a marked reduction in cell viability. At a concentration of 2.5 µM, treatment with these compounds resulted in mean viability values below 25%. In contrast, non-malignant HEK293 cells maintained high viability upon exposure to most of the tested compounds, with more than 75% of compounds displaying viability values above 80% (Figure 6). This differential response indicates selective cytotoxicity toward neuroblastoma cells and provides a rationale for further dose-dependent evaluation of compounds 7a, 7b, and 8aa. Such selectivity is particularly relevant in the context of neuroblastoma, where therapeutic resistance and treatment-related toxicity remain significant challenges. To characterize the cytotoxic potency of the most active candidates, compounds 7a, 7b, and 8aa were evaluated in concentration-dependent studies following 48 h exposure in Kelly and SHSY5Y neuroblastoma cells, as well as in non-malignant HEK293 cells. Half-maximal inhibitory concentrations IC_50_ values were determined by nonlinear regression analysis of WST-1 absorbance data at 450 nm, as shown in Figure 7A–I.

Neuroblastoma Kelly cells (see Figure 7D–F) exhibited pronounced sensitivity to the lead compounds, with IC_50_ values of 1.7 µM for 7a, 1.3 µM for 7b, and 1.5 µM for 8aa. In contrast, SHSY5Y cells (Figure 7G–I) displayed reduced sensitivity, with IC_50_ values ranging from 3.4 to 5.4 µM, indicating variability in response among neuroblastoma models. Such differences may be associated with distinct genetic backgrounds, metabolic profiles, or cellular uptake mechanisms known to differ between neuroblastoma cell lines. Importantly, all three compounds showed low cytotoxicity in non-malignant HEK293 cells (Figure 7A–C), with IC_50_ values exceeding 10 µM. This differential activity profile demonstrates selective cytotoxicity toward neuroblastoma cells and suggests a favorable therapeutic window for compounds 7a, 7b, and 8aa. To benchmark the cytotoxic potency of the lead compounds, their activity was compared with that of doxorubicin, a clinically established chemotherapeutic agent used in the treatment of solid tumors, including neuroblastoma. The IC_50_ values of doxorubicin were determined in both neuroblastoma Kelly cells and non-malignant HEK293 cells under identical experimental conditions (Figure 8A,B).

In Kelly cells, doxorubicin exhibited an IC_50_ of approximately 1.3 µM, comparable to that observed for compound 7b. In contrast, doxorubicin showed pronounced cytotoxicity in non-malignant HEK293 cells, with an IC_50_ of 2.1 µM, indicating limited selectivity toward malignant cells. By comparison, compound 7b retained comparable potency in neuroblastoma cells while displaying substantially reduced cytotoxicity in HEK293 cells. This differential activity profile highlights the improved selectivity of compound 7b and suggests a more favorable therapeutic window relative to doxorubicin.

To assess whether the observed cytotoxic activity of compound 7b extends beyond neuroblastoma, its potency was further evaluated across a panel of human cancer cell lines, including hepatocellular carcinoma (HepG2, Huh7), breast cancer (MCF7, MDA-MB-231), cervical cancer (HeLa), melanoma (M21), and neuroblastoma (Kelly, SHSY5Y) (Figure 9).

As shown in Figure 9, compound 7b demonstrated micromolar IC_50_ values (1.3–9.1 µM) across all cancer lines tested, with the most pronounced sensitivity observed in Kelly, SHSY5Y and M21 cells. This broad cytotoxic activity suggests that compound 7b may interfere with fundamental cellular pathways such as cell cycle regulation or apoptotic signaling that are commonly dysregulated across multiple tumor types. These findings indicate that the anticancer activity of compound 7b is not restricted to neuroblastoma and may involve shared regulatory pathways across multiple tumor types.

2.3. Selectivity Index

To assess the therapeutic window and tumor-targeting selectivity of the synthesized oxazole derivatives, the selectivity index (SI) was calculated as the ratio of IC_50_ values in non-malignant HEK293 cells to those in malignant neuroblastoma cells (Kelly and SHSY5Y). Compounds 7a, 7b, and 8aa demonstrated pronounced selectivity toward neuroblastoma cells. Specifically, compound 7b exhibited an IC_50_ of approximately 1.3 µM in Kelly cells and >10 µM in HEK293 cells, yielding an SI > 7.7. Compound 7a showed an IC_50_ of ~1.7 µM in Kelly cells and >10 µM in HEK293 cells, corresponding to an SI > 5.9, while compound 8aa displayed an IC_50_ of ~1.5 µM in Kelly cells and >10 µM in HEK293 cells, resulting in an SI > 6.7. In contrast, the reference chemotherapeutic agent doxorubicin exhibited limited selectivity, with IC_50_ values of ~1.3 µM in Kelly cells and ~2.1 µM in HEK293 cells, yielding an SI of approximately 1.6. Overall, these results highlight the markedly improved selectivity profile of compounds 7a, 7b, and 8aa relative to doxorubicin and support their further investigation as neuroblastoma-selective anticancer agents.

2.4. Biodegradability

Active pharmaceutical ingredients (APIs) and their metabolites/transformation products are found as pollutants in the environment, impacting human and environmental health. Designing APIs with high biodegradability is desirable but there is no clear agreement on how to implement these criteria in practice [27].

In the CBT, the term “readily biodegradable” refers to those that pass the test with 60% or higher degradation, indicating they will biodegrade rapidly and completely in aquatic environments under aerobic conditions. The CBT experiments were run for 28 days; each day the oxygen concentration was measured and logged for each of the duplicates in the series. The three compounds 7a, 7b, and 8aa did not show any significant biodegradation as the biodegradation % of all the three compounds after 28 days is only 5–10% and did not pass readily biodegradability test. The graphs are given on Figure S97.

Low biodegradability is caused by either toxicity of the studied compounds to the inoculum bacteria or their persistence under experimental conditions. The oxazoles 7a, 7b, and 8aa demonstrated relatively low toxicity against the healthy cells and did not affect the toxicity control in the CBT (see Figure S97) confirming they remain chemically intact under CBT conditions. Indeed, environmental (bio)degradability helps in designing hit compounds that resist breakdown by enzymes in the human body yet finding a ‘window of opportunity’ where such hits are both metabolically stable enough and environmentally biodegradable remains a challenge [27,28].

2.5. Molecular Modeling

Consistent with our previous investigations [29,30,31] and supported by the experimental results obtained herein, compounds 7a, 7b, and 8aa were selected as lead candidates for molecular docking studies, using three-dimensional structures obtained from the RCSB Protein Data Bank [29], where the following protein crystal structures were used: anaplastic lymphoma kinase in complex with Crizotinib—ALK (PDB ID: 2XP2); cyclin-dependent kinase series CDK2 (PDB ID: 3QXO), CDK4 (PDB ID: 7SJ3), CDK6 (PDB ID: 8I0M), CDK7 (PDB ID: 1UA2), CDK9 (PDB ID: 6Z45), as well as the checkpoint kinase 1—CHK1 (PDB ID: 1ZYS), apoptosis regulator BCL2 (PDB ID: 4LVT), Aurora-A kinase in complex with N-Myc (PDB ID: 5G1X) and poly[ADP-ribose] polymerase 1 PARP1 (PDB ID: 6I8T). First, molecular docking techniques were validated using redocking co-crystallized ligands into the proteins using docking scores from AutoDock Vina v1.2.5 program (reported as estimated binding free energies, ΔG, in kcal/mol). Next, molecular docking of 7a, 7b, and 8aa was done, and the results are presented in Table 1 and Figure 10.

Table 1 shows molecular docking validation by redocking co-crystallized ligands into the active sites of the target proteins: ALK kinase, as well as cyclin-dependent kinases CDK2, CDK4, CDK6, CDK7, CDK9, and CHK1 kinase, BCL2 and Aurora A kinase, and polymerase PARP1. The registered binding energies (∆G) for these protein–ligand interactions varied from −8.2 to −11.5 kcal/mol, and RMSD values were 1.01–2.33 Å. According to the results of molecular docking, the most stable protein–ligand complexes of compounds 7a, 7b, and 8aa were observed for the Aurora A kinase, with values of ∆G ranging from −10.8 to −10.9 kcal/mol. Molecular docking of these compounds into the other proteins did not show high complexation energies ∆G (from −7.2 to −9.1 kcal/mol). Notably, the studied compounds exhibit predicted binding affinities within the ATP-binding sites of several kinases, as reflected by their calculated binding free energies (∆G = −7.9–9.1 kcal/mol) in the ATP-binding centers of cyclin-dependent kinases (CDK2, CDK4, CDK6, CDK7, and CDK9), which is comparable to the redocking energy of ligands (∆G = −8.2–9.3 kcal/mol). The low energy of the complexation was demonstrated by ALK kinase (∆G = −7.1 to −7.5 kcal/mol) and the apoptosis regulator BCL2 (∆G = −7.4 to −8.0 kcal/mol). Thus, the most energetically favorable complexation for the studied compounds was found to occur in the ATP-binding site of the Aurora A kinase, with a ΔG value of −10.8 to −10.9 kcal/mol, which is equal to the binding energy of adenosine 5′-diphosphate (ADP) with ∆G = −10.9 kcal/mol. This observation is consistent with the literature. First-generation Aurora A inhibitors are ATP-competitive molecules, many of which have progressed into different stages of clinical evaluation [31]. Successful piperazine-containing Aurora A kinase inhibitors include ENMD-2076 [32] and VX-680 [33], both evaluated in clinical studies. Next, we studied the molecular docking of compounds 7a, 7b, and 8aa into the ATP-binding site of Aurora A kinase (Figure 11).

Figure 11 illustrates the result of the molecular docking of compounds 7a, 7b, and 8aa into the ATP-binding site of Aurora A kinase. The complexation of compound 7a (Figure 11a,b) with a binding energy of ∆G= −10.8 kcal/mol is stabilized by three hydrogen bonds; two hydrogen bonds formed between the carbonitrile group of compound and amino acids Asp274 (2.68 Å) and Lys162 (2.17 Å), and one bond present between the piperazine moiety and the amino acid residues Lys141 (2.22 Å). This ligand–protein complex is also stabilized by three electrostatic interactions between the compounds’ sulfonyl group, the amino acid Lys162 (4.04Å), and the cofactors Mg^2+^ (4.28–4.39 Å). Notably a metal-acceptor interaction (2.57Å) is formed among the cofactor Mg^2+^ and the compounds’ sulfonyl group. Also, this complex is stabilized by eight hydrophobic bonds (3.95–5.36 Å) with amino acids Ala160, Lys162, Leu139, Val147, and Leu263 with bond distances 3.92–5.08 Å.

The ligand–protein complex of compound 7b (Figure 11c,d) with a binding energy of ∆G= −10.9 kcal/mol is stabilized by eight hydrogen bonds; two H-bonds between the compound carbonitrile group and amino acid residues Asp274 (2.66 Å) and Lys162 (2.31 Å), three H-bonds between the sulfonyl group and the amino acids Lys162 (2.22–3.47 Å), and Gly142 (3.48 Å), and three H-bonds between piperazine ring and amino acids—Lys141 (3.48 Å), Glu260 (2.65Å) and Asn (3.39 Å). Also, a metal–acceptor interaction (3.14 Å) exists between the cofactor Mg^2+^ and the 7b’ sulfonyl group. Furthermore, this ligand–protein complex is stabilized by seven alkyl and Pi-alkyl hydrophobic bonds (3.75–5.43 Å) with amino acids Leu139, Val147, Lys162, and Leu263.

The 8aa ligand-protein complex (Figure 11e,f) is characterized by a binding energy of ∆G= −10.8 kcal/mol and stabilized by eight hydrogen bonds. The first sulfonyl group forms two H-bonds with Lys258 (2.45–2.48 Å), and two H-bonds form between the piperazine ring and amino acids Asp256, Asp274 (3.38–3.42 Å). The second sulfonyl group forms one H-bond with Lys143 (2.74 Å), and the oxazole forms two H-bonds with the amino acid residues Lys143 (3.28 Å) and Lys162 (2.83 Å). The cofactor Mg1392 forms two metal–acceptor interactions with the oxazole ring (2.42 Å) and the second sulfonyl group (2.62 Å). Also, this ligand–protein complex is stabilized by three electrostatic interactions: between the compounds’ first sulfonyl group and Trp277 (4.88 Å); among the oxazole ring and cofactor Mg1391 (4.03 Å), and the amino acid Asp274 (3.70 Å). In addition, the ligand-protein complex is stabilized by three hydrophobic bonds (3.63–4.79 Å) with amino acids Val147 and Leu263. Thus, the molecular docking results of compounds 7a, 7b, and 8aa into the ATP-binding site of human Aurora A kinase are characterized by high binding energy, forming multiple H-bonds, electrostatic and hydrophobic interactions, and forming metal–acceptor interactions of compounds with Mg^2+^ cofactors, investigating the possible anticancer mechanism of action of compounds 7a, 7b, and 8bb by molecular docking demonstrates a binding model similar to the ADP molecule (Figure 12).

Our docking studies with the structure of Aurora-A kinase (PDB ID: 7ZTL) also confirm this interaction model [34]. Next, the pharmacokinetic aspects of compounds 7a, 7b, and 8aa were investigated by ADMET 3.0 web tool [24].

2.6. Integrated Biology and Docking Discussion

The strong binding of the compounds 7a, 7b, and 8aa with the ATP-binding site of the Aurora A kinase, with a ΔG value of −10.8 to −10.9 kcal/mol (ADP ∆G = −10.9 kcal/mol), is in good agreement with the literature data. First-generation Aurora A inhibitors are known to act as ATP-competitive agents that occupy the ATP-binding pocket, and many representatives of this class have progressed to various stages of clinical evaluation [35]. A successful example is the piperazine-containing Aurora A kinase inhibitor (IC_50_ = 1.86 nM) ENMD-2076, 6-(4-methylpiperazin-1-yl)-N-(5-methyl-1H-pyrazol-3-yl)-2-[(E)-2-phenylethenyl]pyrimidin-4-amine, which is currently in Phase I/II clinical trials [36]. In the literature, a selective Aurora A inhibitor (IC_50_ = 0.6 nM), Tozasertib (VX-680), N-[4-[4-(4-methylpiperazin-1-yl)-6-[(5-methyl-1H-pyrazol-3-yl)amino]pyrimidin-2-yl]sulfanylphenyl]cyclopropanecarboxamide, is described [37].

It is also known that Aurora A kinase inhibitors bind to the ATP site and cause significant conformational changes in the kinase structure, preventing binding to oncoprotein N-Myc, to block the formation of the Aurora A/N-Myc complex. The formation of this complex plays a crucial role in stabilizing the oncoprotein N-Myc, as key transcription factor that regulates cell cycle progression, cell proliferation, and the metastasis of neuroblastoma. Given the more favorable (i.e., more negative) binding energies compared to the reference in our study, Aurora A kinase represents a plausible molecular target for these compounds. In our opinion, the high activity of compounds 7a, 7b, and 8aa against neuroblastoma cells Kelly expressing N-MYC is associated with their high energy of complex formation in the ATP-binding center of Aurora A kinase, disruption of its conformation, and, as a consequence, destabilization of the oncoprotein N-Myc with antimitotic and antiproliferative action. This hypothesis is supported by the literature data, indicating the development of specific methods for targeting Aurora A in cancer treatment, which initially concentrated on designing various Aurora A kinase inhibitors and are currently in clinical trials [38]. Nevertheless, this proposed mechanism requires direct biochemical validation, such as kinase activity assays or assessment of N-Myc protein stability.

In contrast to Kelly cells, SHSY5Y neuroblastoma cells do not harbor MYCN amplification and are therefore less dependent on Aurora A-mediated stabilization of the MYCN protein. Despite this, SHSY5Y cells remained sensitive to treatment with compounds 7a, 7b, and 8aa, albeit with moderately higher IC_50_ values (3.4–5.4 µM), suggesting the involvement of additional molecular pathways. Notably, SHSY5Y cells have previously been shown to respond to pharmacological inhibition of cyclin-dependent kinases (CDKs). For example, the selective CDK4/6 inhibitor Palbociclib induces G1 phase arrest and suppresses proliferation in neuroblastoma models, including SHSY5Y [39]. Moreover, CDK7 and CDK9—key regulators of transcriptional elongation via phosphorylation of RNA polymerase II—have emerged as therapeutic vulnerabilities, particularly in MYCN-low or MYCN-negative neuroblastoma subtypes [40]. In agreement with this, our docking results indicate favorable predicted binding affinities of compounds 7a, 7b, and 8aa toward multiple CDKs, including CDK4 and CDK7 (ΔG = −9.0 to −9.1 kcal/mol), suggesting that inhibition of CDK-driven cell cycle and transcriptional programs may contribute to the observed cytotoxic effects, particularly in SHSY5Y cells.

Collectively, these findings support a broader multi-target activity profile for compounds 7a–8aa, in which cytotoxicity may arise from the concurrent disruption of mitotic regulation via Aurora A and transcriptional or cell-cycle control via CDKs. Such multitargeted mechanisms are increasingly recognized as advantageous in anticancer drug development, as they may enhance therapeutic efficacy while reducing the likelihood of resistance, particularly in aggressive pediatric malignancies [41]. Consistent with this interpretation, the lead compound 7b also exhibited cytotoxic activity across several non-neuroblastoma cancer cell lines, including M21 (melanoma), MCF7 (breast cancer), and Huh7 (hepatocellular carcinoma), with IC_50_ values in the low micromolar range. This pattern is characteristic of cytotoxic agents that interfere with fundamental processes such as cell cycle progression, mitotic spindle formation, or transcriptional regulation, which are commonly dysregulated in rapidly proliferating tumors [42].

Altogether, these results suggest that the therapeutic potential of the investigated oxazole derivatives may extend beyond neuroblastoma and warrants further evaluation in other high-proliferation tumor types. However, comprehensive mechanistic validation—including kinase activity assays, cell-cycle analysis by flow cytometry, and transcriptomic profiling—will be required to precisely delineate the relative contributions of Aurora A and CDK inhibition to the observed anticancer effects [43].

2.7. ADMET Evaluation

Table 2 presents the significant ADMET properties of compounds 7a, 7b, 8aa, and doxorubicin (to compare) as calculated using the online ADMETlab 3.0 web server.

The synthesized compounds exhibited acceptable physicochemical properties, including a rotatable bond count of (3–4) and a count of hydrogen bond donors (0–1) and acceptors (7–9). The topological polar surface area of compounds 7a, 7b, and 8aa is 127–157 Å^2^, which is slightly lower than the doxorubicin value (222 Å^2^). The coefficient lipophilicity (logP) of compounds is in the acceptable range (0–3 log mol/L) and is 1.424 for compound 7a and 1.208 for doxorubicin. The compounds 7b and 8aa exhibit higher lipophilicity, with values of 2.013 and 2.048, respectively. The values of BBB permeability of compounds 7a (−0.822) and 7b (−0.87) are below the permissible threshold (log BB < −1), and that of compound 8aa (−1.474) is slightly below the threshold and equal to the corresponding value of doxorubicin (−1.379). The values of CNS permeability for all are <3 and are considered unable to penetrate the CNS. The absorption results show that compounds 7a and 7b are well permeable through intestinal cell membranes (logCaco-2 > 5.15), and compound 8aa has moderate permeability (logCaco-2 = −4.815). Compounds 7a and 7b do not interact with P-glycoprotein as a biological barrier; however, compound 8aa might be a P-glycoprotein inhibitor. The excretion of all compounds is estimated by plasma clearance, which ranges from 5.762 to 6.382 mL/min/kg, indicating moderate clearance (5–15 mL/min/kg). The doxorubicin plasma clearance is 2 times higher, amounting to 14.244 mL/min/kg. Additionally, the results of the compounds’ half-lives are obtained, which ranges from 0.686 to 0.821 h, representing an ultra-short half-life value (<1 h). Doxorubicin has a value of 3.774 h and is a short half-life drug. The toxicology properties of compounds 7a, 7b, and 8aa are characterized by positive hepatotoxicity (similar to doxorubicin) and a low tolerated dose log (mg/kg/day) from −0.158 to −0.554; this dose for doxorubicin is higher (0.081). The values of the rat oral acute toxic doses (LD_50_) of compounds 7a, 7b, and 8aa are comparable to those of doxorubicin (2.4–2.5 mol/kg). Additionally, Figure 13 illustrates the ADMET characteristics of compounds 7a, 7b, 8aa and doxorubicin as estimated by the ADMETlab 3.0 web server.

3. Experimental Section

3.1. Chemicals Materials and Methods

All chemicals were used without further purification in the synthetic procedures, including doxorubicin hydrochloride (batch WRS DR 02, was provided by Gemini PharmChem Mannheim GmbH/Synbias Pharma AG), commercially available reagents (Sigma-Aldrich, Taufkirchen, Germany), reagent-grade solvents (Lach-Ner, Neratovice, Czech Republic), and deutero solvents (Eurisotop, Saint-Aubin, France). The reaction progress was monitored using thin-layer chromatography (TLC) on silica gel 60 F254 plates (Merck KGaA, Darmstadt, Germany). Melting points (M.p.) were determined in open capillary tubes using a Stuart SMP40 apparatus with a 2 °C/min ramp, and values are reported in °C. NMR spectra (^1^H and ^13^C) and 2D HSQC plots were recorded on a Bruker Avance III 400 MHz spectrometer in DMSO-d_6_, with residual SO(CD_3_)(CD_2_H) (δH = 2.50 ppm) or SO(CD_3_)2 (δ_C_ = 39.52 ppm) as the internal standard. HPLS analysis was conducted on an Agilent 6540 UHD Accurate-Mass Q-TOF LC/MS G6540A Mass Spectrometer. Elemental analysis (CHNS) was performed using a Vario Micro Cube device. Column chromatography was performed using Macherey-Nagel Silica 60, 0.04–0.063 mm silica gel. FT-IR spectra were recorded on a Shimadzu IRTracer-100 spectrometer (Kyoto, Japan) using KBr pellets (1% w/w), with a resolution of 2 cm^−1^ and detection at 254 nm. Dichloroacrylonitriles (1a–1c) were prepared as described in the literature [26,44,45]. The starting 4-cyano-2-aryloxazole-5-sulfonyl chlorides (see Scheme 2, structure IV) were synthesized according to a literature procedure [46]; structures 7a and 7b first reported in [47] were confirmed in this work. The synthetic procedures and spectral data of all compounds are collected in the Supplementary Information (Figures S1–S96).

3.2. Biology: In Vitro Anticancer Studies

Cell cultures. To evaluate the cytotoxicity and anticancer potential of the synthesized oxazole derivatives, a panel of human cancer cell lines—including HeLa (cervical carcinoma), HepG2 and Huh7 (hepatocellular carcinoma), MDA-MB-231 and MCF7 (breast adenocarcinoma), M21 (melanoma), and Kelly and SHSY5Y (neuroblastoma)—was employed. These cell lines were obtained from the American Type Culture Collection. Additionally, human embryonic kidney cells HEK293 cells were used as a non-cancerous control to assess selectivity toward malignant cells. Cells were cultured in DMEM supplemented with 10% bovine calf serum and 5% penicillin/streptomycin, and maintained under standard conditions (37 °C, 5% CO_2_). Treatments were applied 24 h after seeding, followed by a 48 h incubation with either vehicle or compound-containing medium. Cell viability was assessed using the WST-1 assay, which quantifies mitochondrial activity through tetrazolium salt reduction by measuring IC_50_, or the half maximal inhibitory concentration in this test against neuroblastoma cancer cell, we studied drug activity and possible cellular pathways of action. Additionally, to learn relative efficacy and SI of the hit compounds (7a, 7b and 8aa), as indicated by IC_50_, doxorubicin was used in identical experiments, especially in targeting specific tumor types. The cells were propagated in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco), supplemented with 10% bovine calf serum (Gibco) and 5% penicillin/streptomycin. All cell lines were incubated at 37 °C in a humidified 5% CO_2_ and 95% air atmosphere.

Cell treatment procedures. The cells were plated at a density of 2.5 × 105 cells/well (The Countess Automated Cell Counter, Invitrogen, Bothell, WA, USA.) in 96-well plates and incubated overnight. After 24 h of incubation, 100 µL of either fresh media or fresh media containing diluted agents was added into each well and incubated for a further 48 h. The experiments with addition of MeOH were used as solvent control.

Cell viability measured by WST-1. The effects of agents on the viability of cells were determined using the cell viability assay WST-1 (Roche). WST-1 allows colorimetric measurement of cell viability due to reduction in tetrazolium salts to water-soluble formazan by viable cells. The amount of formed formazan dye correlates with the number of viable cells. The measurements were completed 48 h after the cell treatments. The experiments, with an addition of 5 µL of MeOH, were used as a solvent control. Five microliters per well of the WST-1 reagent was added to 100 µL of the cell culture medium, incubated at 37 °C for 2 h, after which the absorbance was measured at 450 nm by using a GENios Pro Microplate Reader (Tecan Group Ltd., Grödig, Austria). WST-1 reduction correlates with the number of metabolically active, viable cells, thus providing a quantitative readout of cytotoxicity.

Statistical analysis was performed using two-way Anova, together with Dunnett’s multiple comparisons test. The graphs represent data from at least 3 independent experiments, all performed in triplicate, as the mean ± standard deviation. In the cell viability assay, data of treated with solvent cells showed as control cells. A statistical significance between control and cells treated with compounds showed, where p < 0.05 is represented as *, p < 0.01 as **, p < 0.001 as *** and p < 0.0001 as ****. Statistical analysis was performed with GraphPad Prism 9.

3.3. Molecular Modeling Methods

Molecular docking was performed using AutoDock Vina v1.2.5 program [48]. The X-ray crystal structures of target proteins complexed with inhibitors were obtained from the RCSB Protein Data Bank (PDB). Prior to docking, protein structures were prepared using AutoDockTools (ADT) v.4.2 software [49], which provides a graphical user interface for model setup. Polar hydrogen atoms were added, all atoms were renumbered, and Gasteiger partial charges were assigned using ADT. The prepared protein structures were then saved in PDBQT format for use in docking simulations. The 2D and 3D structures of the ligands 7a, 7b, and 8aa were generated and refined through pre-optimization using the ChemAxon MarvinSketch v.23.11.0 software [50]. The ligand structure optimization and its energy were minimized by the Avogadro v.1.2.0 program [51]. This procedure was performed using the Auto Optimize tool, which employed molecular mechanics calculations to refine the molecular geometry by minimizing the potential energy. We used the force field with MMFF94s, the “steepest descent” algorithm, and the default setting for “Steps per Update” of 4. Next, the three-dimensional structures of the ligands were prepared for docking studies using the AutoDockTools program and saved in PDBQT format for subsequent molecular docking. Docking studies were performed using AutoDock Vina with a grid spacing of 0.375 Å and a grid map ranging from 30 × 30 × 30 Å to 40 × 40 × 40 Å. The docking centers were the geometric centers of the co-crystallized ligands. Under these conditions, the optimized protein and ligand structures served as inputs for docking simulations targeting the defined active site. The AutoDock Vina scoring function was employed to evaluate and rank the docking poses based on their predicted binding affinities. The docking output files were rendered and examined for key interactions between the ligands and the amino acid residues constituting the active sites using BIOVIA Discovery Studio Visualizer 2019 [52]. To ensure the reliability of the docking results, five to six independent runs were performed, yielding up to nine distinct docking poses. Pose selection was based primarily on the AutoDock Vina-predicted binding affinities (kcal/mol) and RMSD values. Additionally, the presence and geometry of potential hydrogen bonds and electrostatic interactions were considered.

3.4. ADMET

The online tool, ADMETlab 3.0 web server [53], was used to calculate the ADMET properties of compounds 7a, 7b, and 8aa in comparison to Doxorubicin [54]. ADMETlab 3.0 tool includes a multi-task DMPNN (directed message passing neural network) architecture based on molecular descriptors. This approach ensures high-speed calculation and performance for each endpoint, along with an elevated level of accuracy and robustness. Additionally, the pkCSM web server was used to predict some ADMET properties [55].

3.5. Biodegradability Study

Biodegradability of the selected oxazoles 7a, 7b, and 8aa was determined using modified aerobic biodegradation test OECD 301D [56] known as Closed Bottle Test (CBT) [57], usually implied as an initial screening test for organic compounds [58].

CBT setup with modifications where biological oxygen consumption is measured with an optode oxygen sensor system using PTFE-lined PSt3 oxygen sensor spots (Fibox 3 PreSens, Regensburg, Germany), allows measuring BOD without dispensing it from the stock solution each time.

Each CBT run consisted of four different series, each done in duplicates. First was “reference series” in which readily biodegradable sodium acetate in a known concentration (6.41 mg/L) was added to a flask of mineral medium inoculated with effluent from a wastewater treatment plant. As sodium acetate is rapidly biodegradable it acted as a reference and control for monitoring activity of microbes in the inoculum [59]. In the test series a studied compound as a sole source of carbon was added to the inoculated mineral medium. The test compound was added in a concentration corresponding to theoretical oxygen demand (ThOD) of approximately 5 mg/L. The details of the calculated ThOD and the amount of test substance used to measure biodegradability in the CBT are listed in Table S1.

Effluent from wastewater treatment plant was collected at municipal wastewater treatment plant in Tallinn, Estonia (Paljassaare wastewater treatment plant, 59°27′55.5″ N 24°42′08.8″ E). Results from each run were accepted if the following criteria were met: (i) the difference in extremes of replicate values at the plateau is less than 20%, (ii) oxygen concentration in test series bottles must not fall below 0.5 mg/L at any time, and (iii) sodium acetate in reference series must be degraded ≥60% by day 14. Blank bottle oxygen consumption was also monitored to avoid the possibility of the system turning from aerobic to anaerobic.

4. Conclusions

Thirty-two 5-piperazine-containing 1,3-oxazole-4-carbonitriles were rationally designed and synthesized in yields ranging from 68% to 83%, and their structures were confirmed by ^1^H and ^13^C NMR spectroscopy, HPLC, and elemental analysis. In vitro cytotoxicity assays (WST-1) identified compounds 7a, 7b, and 8aa as lead candidates, demonstrating reproducible micromolar activity against neuroblastoma cell lines (Kelly and SHSY5Y) while maintaining low toxicity toward non-malignant HEK293 cells (IC_50_ > 10 µM), resulting in favorable selectivity indices (SI > 5.9). Notably, compound 7b displayed potency comparable to doxorubicin in Kelly cells but with substantially reduced cytotoxicity in HEK293 cells. Across a broader cancer panel, 7b retained low-micromolar IC_50_ values, indicating activity in rapidly proliferating tumor cell types.

In silico molecular docking suggested favorable binding of compounds 7a, 7b, and 8aa to the ATP-binding site of Aurora A kinase in complex with N-MYCN, with calculated binding free energies comparable to that of ADP, supporting Aurora A as a plausible molecular target. Additional predicted interactions with cyclin-dependent kinases (CDK4 and CDK7) raise the possibility of a multi-target mechanism involving both mitotic and transcriptional or cell-cycle-associated pathways.

Predicted ADMET profiles indicated acceptable physicochemical properties and moderate clearance, although potential hepatotoxicity comparable to doxorubicin and limited biodegradability under OECD 301D conditions were observed, reflecting common trade-offs at the hit-to-lead stage. Overall, compounds 7a, 7b, and 8aa represent promising lead structures for further preclinical investigation against neuroblastoma, with potential relevance for other rapidly proliferating solid tumors. The favorable in vitro selectivity and predicted multi-target binding profile provide a strong rationale for continued optimization and targeted mechanistic validation.