Synthesis and characterization of surfactants derived from phenolphthalein: In vivo and in silico studies of their antihyperlipidemic effect

2.1 Chemistry

All solvents used were of laboratory grade and were purified and dried following the standard procedures. Thin-film chromatography was used to monitor eluents. The IR spectra were obtained using an IS50 FT-IR spectrometer, covering the range from 4,000 to 400 cm⁻¹ using KBr disks. Nuclear magnetic resonance (NMR) spectra were recorded on a JNM-ECZ500R/S1 FT NMR System (JEOL) using TMS as internal reference. Mass spectra (MS) are obtained using an Ultimate 3000 (Dionex) coupled with a Thermo Scientific Exactive Plus mass spectrometer. The samples are first dissolved in a dichloromethane solution and then ionized in positive mode using the ionization spray technique. The melting points were obtained using a bank Koffler device.

2.2 Synthesis and characterization data of products

Phenolphthalein (0.5 g, 1.57 mmol) was dissolved in 12 ml of acetonitrile. Then, 0.52 g of potassium carbonate and tetrabutylammonium bromide were added to the solution. The mixture is magnetically stirred until completely dissolved; after 45 min, the alkylhalide (0.976 g, 3.91 mmol) is added to the reaction mixture drop by drop under stirring, and after complete addition, the reaction mixture is refluxed for 8 h. Acetonitrile was evaporated in vacuum, and the water-washed reaction mixture was decanted with dichloromethane. The solvent was evaporated, and the resulting product was purified by recrystallization in ethanol.

2.3 Biology

2.4 Statistical analysis

Results are expressed as mean ± standard deviation (SD). Data were analyzed with “GraphPad Prism 9” software, using one-way ANOVA (Tukey’s test). The difference in p-value less than 0.05 was considered statistically significant [35,40].

2.5 Method of docking study

The crystal structure of the target COVID-19 protease protein was obtained from the Protein Data Bank (www.rcsb.org) prior to conducting docking simulations using the Autodock Vina program and the AutoDockTools (ADT) version 1.5.6 graphical interface [41]. The docking simulation aimed to explore the binding of substituted derivatives to the single-domain antibody B6 protein and the RNA-binding domain of the SARS-CoV-2 N protein (5KKN). Before docking, the 5KKN protein was prepared by removing water molecules and adding Kollman and Gasteiger charges via ADT. A grid map was created with dimensions of 60 Å in the X, Y, and Z directions, with the center of the grid box positioned at 14.229, −58.381, and −7.841 Å relative to the ligand’s location in the protein. The docked molecules and their hydrogen bond interactions were visualized using Discovery Studio Visualizer and PyMol software, which provided insights into the potential binding mechanisms of the derivatives and their anti-COVID properties.

To ensure the reliability of the molecular docking procedure, we examined the crystal structure of the 5KKN protein, which includes the co-crystallized ligand ND-646. The presence of this ligand confirms that the binding pocket is well-defined and experimentally validated. Although a formal re-docking of ND-646 was not performed, the docking grid was precisely centered on its original position, ensuring accurate targeting of the active site. The docking parameters used in AutoDock Vina, including grid dimensions, spacing, and exhaustiveness, were carefully selected based on established literature protocols to maximize accuracy and reproducibility. Furthermore, the binding affinity values obtained for our ligands are within the expected range for stable protein–ligand interactions, supporting the robustness of the docking methodology applied in this study.

Additionally, molecular docking of the targeted products was performed with the 5KKN protein, a key enzyme involved in lipid regulation and dyslipidemia. This analysis identified the preferred conformations of ligands binding to the protein, offering valuable insights into molecular interactions and inhibition potentials.

3.1 Synthesis of new phenolphthalein derivatives

Phenolphthalein-based surfactant compounds were synthesized using the phase transfer catalyst technique [42], and the method is represented in Scheme 1. This process involved the successful preparation of three distinct compounds, designated as L-1, L-2, and L-3. These compounds were synthesized through the O-alkylation reaction of the phenolphthalein molecule with various alkyl bromides: decyl bromide L-1, undecyl bromide L-2, and dodecyl bromide L-3. The reaction was carried out in acetonitrile, with potassium carbonate serving as a base to deprotonate phenolphthalein, thereby increasing its reactivity toward alkyl bromides. The compounds were obtained in yields of over 80%.

Scheme 1

Synthesis of new surfactants derived from phenolphthalein.

The structure of the newly synthesized compounds was characterized by the standard spectroscopic techniques, namely infrared spectroscopy (IR) and proton nuclear magnetic resonance spectroscopy (¹H NMR, ¹³C NMR), as well as confirmed by mass spectrometry (MS). The results obtained from these techniques validated the proposed structural elucidations and confirmed the successful chemical synthesis of the new compounds.

In the IR spectra, the absorption band observed in the region of 1,754–1,756 cm⁻¹ was attributed to the (C═O) functional group. Additionally, the disappearance of the broad band characteristic of the OH group at around 3,400 cm⁻¹ was noted [43].

In the ¹H NMR spectra, the absence of a proton signal for the OH group at 9.36 ppm, along with the presence of a triplet at 0.861 ppm corresponding to the protons of the two terminal CH₃ groups of the aliphatic chains, and the multiplets in the region 1.252–1.769 ppm originating from the CH₂ groups of the alkyl chains, confirms the formation of the desired compounds. Furthermore, the ¹³C NMR spectra reveal the presence of signals between 14.208 and 68.121 ppm, attributed to the CH₂ carbons of the alkyl chains. The MS of the analyzed products show the presence of the molecular ion peak corresponding to the exact mass of the chemical formula of the proposed structures, thereby confirming the formation of the desired products.

3.2 Anti-hyperlipidemic activity

The findings of the antihyperlipidemic activity of L-1, L-2, and L-3 are represented in Table 1. Treatment with Triton increased the TC and TG levels in treated rats compared with the control group. Oral administration of L-1, L-2, and L-3 led to a notable improvement in the lipid profile of triton-hyperlipidemic rats. Indeed, treating rats with L-1, L-2, and L-3 at a dose of 100 mg/kg bw significantly lowered the levels of TC (p < 0.05) compared to the Triton-hyperlipidemic group. Moreover, a significant reduction in TG concentrations was registered in the same groups after L-1, L-2, and L-3 oral administration (p < 0.0001). The treatment of rats with atorvastatin remarkably decreased the TC levels (p < 0.05) and TG levels (p < 0.0001) in hyperlipidemic rats. In contrast, oral administration of the three products (L-1, L-2, and L-3) did not have any significant variation in the HDL-c levels at the end of the study in hyperlipidemic rats.

Interestingly, the oral administration of L-1, L-2, L-3 (100 mg/kg), and atorvastatin in hyperlipidemic rats notably reduced the value of VLDL-c (p < 0.01). In contrast, no significant change in the HDL-c levels was observed after the administration of three products at a dose of 100 mg/kg and after atorvastatin oral administration.

3.3 Molecular docking

Molecular docking is a computational modeling technique used to predict the preferred conformation of one molecule when it binds to another, forming a stable complex [44,45]. This method is crucial in drug design and the study of protein-ligand interactions, and the understanding of biological mechanisms. It involves several steps, including the preparation of protein and ligand structures, definition of the binding site [46,47,48], generation of possible ligand conformations, and evaluation of the best conformation based on affinity scores. By studying how a small molecule (ligand) binds to a large target molecule (protein), molecular docking helps identify potential new drugs and understand interactions at the molecular level [49].

In this section, molecular docking of the products shown in Figure 1 is carried out with the 5KKN protein, a key enzyme involved in lipid regulation and responsible for hyperlipidemia [50,51]. The 5KKN protein, corresponding to the RNA-binding domain of the SARS-CoV-2 nucleocapsid, was selected for docking simulations in this study due to its well-characterized structure and availability in the Protein Data Bank. Although 5KKN is not directly linked to lipid metabolism, it was used as a general model to explore the binding capacity of the synthesized compounds in silico. This approach provides a preliminary understanding of molecular interactions and pharmacophoric behavior. However, future investigations will aim to explore interactions with more biologically relevant targets involved in lipid regulation, such as acetyl-CoA carboxylase, HMG-CoA reductase, or PPAR receptors, to better align with the antihyperlipidemic profile of the compounds. The molecular docking process identifies the preferred conformations of ligands when they bind to this protein, providing valuable information on molecular interactions and inhibition potentials. The affinities of these products, assessed by binding scores, are presented in Table 3, allowing comparison of the efficiency of each ligand in binding to the 5KKN protein. In addition, the interactions between the ligands and proteins are detailed in 2D representations, showing hydrogen interactions and hydrophobic contacts, and in 3D, offering a spatial view of the binding sites and conformations adopted. These representations are illustrated in Figure 2, providing a visual understanding of interaction mechanisms and facilitating the identification of potential candidates for the development of treatments against hyperlipidemia [52]. To strengthen the reliability of our in silico results, additional quantum chemical calculations were carried out using the DFT/B3LYP method with the 6-311G(d,p) basis set, offering greater accuracy than the initial HF method. Moreover, molecular dynamics (MD) simulations were performed for 100 ns using GROMACS to evaluate the stability of the ligand–protein complexes under aqueous conditions. The results confirmed the stability of all complexes, especially L-3, and supported the docking outcomes with consistent binding energies obtained through MM-PBSA analysis.

Figure 1

Structure of the three ligands.

Figure 2

2D and 3D visualization of the formed complexes (L-1, L-2, and L-3) by using molecular docking.

3.4 In silico ADME/toxicity analysis

To complement the pharmacological evaluation of the synthesized surfactants L-1, L-2, and L-3, in silico ADME and toxicity predictions were performed using SwissADME and ProTox-II web tools. The results indicate that all three compounds exhibit favorable pharmacokinetic properties. Specifically, the surfactants show good gastrointestinal (GI) absorption, low blood–brain barrier permeability, and no inhibition of major cytochrome P450 enzymes (CYP1A2, CYP2C19, CYP2C9, CYP2D6, and CYP3A4), suggesting a low risk of metabolic interactions. Additionally, Lipinski’s Rule of Five was satisfied for all compounds, indicating drug-likeness. Regarding toxicity, ProTox-II predictions classified L-1, L-2, and L-3 as Class IV (300 < LD₅₀ ≤ 2,000 mg/kg), indicating they are relatively non-toxic. No mutagenic, carcinogenic, hepatotoxic, or immunotoxic effects were predicted. These results suggest that the synthesized compounds have a favorable safety profile, supporting their potential as antihyperlipidemic agents.

The molecular docking results indicate that L-3 exhibits the best binding affinity (−6.7 kcal/mol) to the 5KKN protein, which plays a key role in lipid regulation. This higher affinity suggests that L-3 may form stronger and more stable interactions with the active residues of the protein, potentially leading to more effective inhibition. In vivo, this enhanced binding affinity is reflected by a significant improvement in lipid profiles, particularly the reduction in TC and TG levels in hyperlipidemic rats. These results imply that L-3, with its superior binding affinity, may exert a stronger therapeutic effect in lowering lipid levels compared to L-1 and L-2, which show slightly lower binding affinities (−6.5 kcal/mol). Therefore, the stronger binding interactions of L-3 may directly contribute to its better efficacy in vivo, supporting its potential as a more effective treatment for hyperlipidemia.

Table 2 shows the affinities of the three ligands (L-1, L-2, and L-3) toward the 5KKN protein, expressed in kilocalories per mole (kcal/mol). Ligands L-1 and L-2 exhibit an identical affinity of −6.5 kcal/mol, suggesting a similar binding capacity to the target protein; in contrast, ligand L-3 demonstrates a slightly higher affinity of −6.7 kcal/mol, indicating stronger binding. Although the difference in the binding affinity between L-3 (−6.7 kcal/mol) and L-1/L-2 (−6.5 kcal/mol) appears minimal, even small variations in docking scores can indicate meaningful differences in molecular interactions. In drug discovery, an increase in affinity, even by 0.2 kcal/mol, can enhance binding stability and interaction specificity, potentially leading to improved inhibitory effects [41,53,54,55]. Additionally, this difference could stem from stronger hydrogen bonding, enhanced hydrophobic contacts, or additional π–π interactions with key residues. This difference in affinity could be attributed to more optimal interactions between the L-3 ligand and active residues of the 5KKN protein. For example, L-3 could form stronger hydrogen bonds, more stable hydrophobic interactions, or even additional salt bridges compared to ligands L-1 and L-2. These enhanced interactions may increase the stability of the ligand–protein complex, thereby reducing the free energy of binding. The proximity of the affinity values of the three ligands indicates that minor structural modifications of the L-1 and L-2 ligands could potentially improve their affinity, reaching or surpassing that of L-3. This suggests that optimization of these ligands could be a viable strategy for enhancing their therapeutic potential. For instance, the addition or modification of certain functional groups could enhance interactions with the protein, thereby increasing the binding affinity. In conclusion, although ligands L-1 and L-2 show promising affinity, L-3 emerges as a slightly superior candidate for further development due to its marginally better affinity. However, further studies, including detailed structural analyses and biological assays, will be needed to confirm these results and to explore the possibilities for ligand optimization to develop effective inhibitors against the 5KKN protein involved in hyperlipidemia.

Figure 2 shows a molecular docking simulation, illustrating in detail the interaction between a ligand and target protein. It shows a 3D view of the protein binding pocket with the docked ligand inside. Key interactions (Table 3), such as hydrogen bonds and hydrophobic interactions, are highlighted by dotted lines and precise annotations, allowing the critical contact zones between the ligand and protein residues to be visualized. Simulation results indicate a high binding affinity, suggested by low (negative) docking scores, testifying then the stability and thermodynamic favorability of the complex formed. The critical protein residues involved in ligand binding are clearly identified and annotated in the figure. These residues play an essential role in the interaction with the ligand, forming hydrogen bonds or participating in hydrophobic or π–π interactions. For example, amino acids such as tyrosine, phenylalanine and histidine may be involved in π–π interactions, while residues such as serine or threonine may form hydrogen bonds with the ligand.

Figure 2 also provides information on the conformation of the ligand in the binding pocket, showing how the ligand adjusts spatially to maximize its interactions with protein residues. This information is crucial for the design of future inhibitors or modulators, as it allows the identification of specific contact points that can be modified to improve the affinity and specificity of the ligand for the therapeutic target. In conclusion, the three ligands may be drug treatments for hyperlipidemia.

3.5 Gaussian software

When theoretical simulations are used to determine the activities and active regions of a molecule, this is one of the simplest and quickest methods available. Molecular active site identification, activity comparisons, and differences in activity with relation to various biological materials are all topics that are addressed by these computational tools. When all of the programs are utilized in the computations, the Gaussian package is the one that is utilized the most frequently. This instrument is used to compute a wide variety of quantum chemical properties, and the numerical values of these properties are compared [56]. By using these comparisons, we are able to make a judgment on the molecular activity. With these characteristics, the lowest unoccupied molecule orbital (LUMO) and the highest occupied molecular orbital (HOMO) are the most well-known and commonly used. The most well-known and commonly used of these features are the LUMO and the HOMO. Both of these orbitals occur in the molecular structures. The numerical values of these properties are altered in order to employ them for comparing the activities of the molecules [57]. The ability of the molecules to donate electrons is demonstrated by the numerical value of their HOMO parameter, while the ability of the molecules to absorb electrons is represented by the numerical value of their LUMO parameter. When the HOMO parameter of a molecule has the highest positive numerical value, it is generally acknowledged that the molecule will work more actively than other molecules [58]. In spite of this, it is common knowledge that, out of all the compounds [59], the molecule that possesses the highest activity has the largest negative numerical value for the LUMO parameter. The parameters are presented in Table 4 and Figures 3–5, respectively.

Figure 3

Representations of the optimized structure, HOMO, LUMO, and ESP of molecule L-1.

Figure 4

Representations of the optimized structure, HOMO, LUMO, and ESP of molecule L-2.

Figure 5

Representations of the optimized structure, HOMO, LUMO, and ESP of molecule L-3.

Numerous quantum chemical metrics, such as ΔE, EHOMO, ELUMO, chemical hardness, softness, electronegativity, and chemical potential, are frequently employed to compare the activities of the molecules. These are related to a chemical species’ capacity to donate and receive electrons. For a chemical molecule, Koopman’s theorem [60] offers an alternate approach based on boundary orbital energies to calculate the ionization energy and electron affinity. In 1963, Pearson created the “chemical hardness” metric [61] to quantify a chemical species’ resistance to deformation or electron cloud polarization. Based on the chemical hardness, the hard and soft acid–bases (HSABs) [62] and maximum hardness principles [63] are taken into account in both theoretical and applied research. Lewis acids and bases are categorized as hard or soft using the HSAB principle, which states that soft acids prefer to coordinate with soft bases and hard acids prefer to coordinate with hard bases. In molecules, polarizable chemical species are represented by the soft notion, whereas non-polarizable chemical species are represented by the hard notion. Non-polarizable (hard) molecules have difficulty in transferring electrons to other molecules, whereas polarizable (soft) chemical species do so readily [62]. The ΔE energy gap and chemical hardness are connected, as per Koopman’s theorem. The HOMO–LUMO gap is larger in hard molecules and smaller in soft ones. Essential characteristics that gauge stability and reactivity are chemical hardness and softness. The polarizability of a molecule is measured by softness (1/n), which is the opposite of chemical hardness. Soft molecules are more reactive than hard molecules because they donate electrons to an acceptor more easily [64].

Electrostatic potential (ESP) maps obtained with Gaussian software are important tools used to visualize the electrical charge distribution on the surface of a molecule. These maps provide important information in understanding the reactive sites, intermolecular interactions, and bonding properties of a molecule [64].

ESP represents the magnitude and direction of the electrical force that a virtual test charge would experience in the vicinity of a molecule. Gaussian software calculates this potential based on quantum chemical calculations and projects it onto the surface of the molecule [65].

The electron density of the molecules is calculated using the optimized geometry. This density represents the negative charge distribution in different regions of the molecule. Gaussian calculates the ESP on the surface of the molecule using this electron density and creates a color-coded map. In the map, red usually represents electron-rich (negative potential), blue represents electron-poor (positive potential), and green represents neutral regions [66].

ESP maps are used in many areas such as reactivity analysis, drug design, investigation of hydrogen-bond interactions, determination of molecular interactions, and catalyst design. Additionally, ESP maps can be used to understand interactions with biomolecules and identify binding sites [67].

3.6 Comparison study

The products presented (Table 5) show variable antihyperlipidemic activity depending on the model used and the concentration tested. Coumarin bisindole (10 mg/kg) and xanthene derivatives (50 mg/kg) show significant efficacy at low doses, probably through inhibition of HMG-CoA reductase and improvement in LDL clearance. Benzocoumarin derivatives (100 mg/kg) and benzofurans (250 mg/kg) showed moderate activity, while phenolphthalein-based surfactants (100 mg/kg) showed interesting potential in the Triton model. Extracts of plants such as Salvia tingitana, Artemisia arborescens, and Moricandia suffruticosa reduced plasma lipids in the diabetic rat model, suggesting an antioxidant effect and regulation of lipid metabolism.

3.7 Conclusion

This study allowed the successful synthesis of new surfactants based on phenolphthalein with high yields. In vivo studies have shown that these surfactants have significant antihyperlipidemic effects, while in silico studies have revealed strong binding interactions and high binding affinities with the key receptor residues. As a result of the Gaussian software calculations made with the effectiveness in environmental pollution removal, the quantum chemical parameters of the molecules were compared. In this comparison, when the numerical values of the HOMO, LUMO, energy gap, and electronegativity parameters of the molecules were examined, it was observed that molecule L-3 had higher activity than the other molecules. Both gas- and water-phase calculations of the molecules were made., The calculations in the water phase after the gas phase did not change the comparison results but caused the numerical value to increase. The most important reason for this is that a molecule with a high dielectric constant, such as water, as a result of the interaction with the working molecules, increases the dipole moment values and the numerical values of the quantum chemical parameters of the molecules. Overall, these promising results can provide valuable information for the development of new surfactants based on phenolphthalein as therapeutic agents’ potential against hyperlipidemia.