Retrieval and preparation of the 3-D structure of STAT3

The RCSB protein data bank was queried to acquire the three-dimensional structure of the target protein STAT3 (PDB ID: 1BG1). Since 1BG1 is a homodimer attached to DNA, the DNA and other heteroatoms were deleted, leaving a clean STAT3 for docking experiments. Energy minimization and geometry optimization were performed using Dock Prep, a built-in tool in UCSF Chimera for creating protein 3D structures. The energy was minimized for the 1000 steepest descent steps using a root mean square gradient of 0.02, and an Amber ff12SB force field, and the prepared protein was saved in .pdb format.

Natural compounds library preparation

ZINC database was utilized to construct a collection of natural compounds, integrating information from the Traditional Chinese Medicine (TCM) database, which encompasses 32,364 commercially accessible natural compounds. This library comprises TCM compounds represented in 3D file formats, specifically .sdf, that accurately depict the processed chemical structures of these compounds.

Virtual screening

VS sorts of compound databases using computer technology and drug design theory-based software. Potentially valuable compounds are screened in large numbers in order to find novel leads for further experimental evaluation. The prepared compound library was screened against the STAT3 using PyRx 0.8 tool [17].

In silico Physicochemical and ADME Prediction

To estimate the physicochemical and ADME properties of the selected compounds, Datawarrior and SwissADME (http://www.swissadme.ch) were employed. These software tools facilitated the assessment and calculation of crucial characteristics of their performance in biological systems and overall pharmacokinetic profiles.

This study used molecular docking-based VS to screen 32,364 TCM compounds for possible hits with significant STAT3 binding scores. The top 20 compounds were initially chosen based on their strong binding affinities with STAT3 (Table 1). The process of filtration to the initial set of top 20 compounds was applied using Lipinski and Veber's tools. As a result, the top six compounds were identified and chosen: ZINC85542844, ZINC4098720, ZINC85543599, ZINC85593523, ZINC85593528, and ZINC85593537. These six compounds passed Lipinski and Veber's rigorous criteria, suggesting their potential as prospective candidates for drug development. Table 2 presents a brief assessment of the physicochemical and drug-likeness values of these compounds. Based on the estimated values, these compounds exhibit a favorable drug-likeness and drug score, suggesting their potential as lead compounds in pharmaceutical development.

An intra-molecular interaction analysis was conducted on the top six compounds. ZINC85593528 interacted with Thr456, His457, Asp237, Asp242, Arg245, Ala241, Lys244, Asn485, Pro487, Gln248, Leu459, Gln247, Trp243, Gln326, Val322, and Val323 residues of STAT3 (Figure 1a); while Glu455, Lys318, His457, Ser319, Thr456, Val322, Phe321, Gln247, Val323, Trp243, Gln248, Arg245, Asn485, Ala241, Leu459, Pro487, Gln326, Ser458, and Lys244 residues of STAT3 interacted with ZINC85593537 (Figure 1b). ZINC85542844 was found to interact with Glu455, Lys318, Ser319, Thr236, Asp237, Ala241, Lys244, Asn485, Asn486, Pro487, His457, Lys488, Leu438, Thr440, Asp371, Lys370, and Ser372 residues of STAT3 (Figure 1c). ZINC4098720 interacted with Pro487, Lys244, His457, Phe321, Thr456, Glu455, Leu454, Lys318, Ser319, Val322, Val323, Trp243, Gln247, Gln326, and Leu459 residues of STAT3 (Figure 1d). ZINC85543599 interacted with Gln247, His457, Lys244, Thr456, Phe321, Glu455, Asp237, Ala241, Arg245, Asn485, Leu459, Pro487, Val322, Val323, Trp243, and Gln326 residues of STAT3 (Figure 1e). Furthermore, ZINC85593523 interacted with Asp237, Glu455, Ser319, Lys318, Val322, Trp243, Gln326, Val323, Gln247, Leu459, Thr456, His457, Pro487, Lys244, Thr236, Ala241, Asn486, and Asn485 residues of STAT3 (Figure 1f).

STX-0119 interacted with Val323, Trp243, Gln326, Gln247, Val322, Leu459, Pro487, Thr236, Ser319, Lys244, Phe321, Asp453, Lys318, Leu454, Glu455, Thr456, and His457 residues of STAT3 (Figure 2).

The superimposition of representative structures revealed that the binding pattern and conformational alignment of hits (ZINC85542844, ZINC4098720, ZINC85543599, ZINC85593523, ZINC85593528, and ZINC85593537) in the STAT3 active site were identical to that of STX-0119 (Figure 3).

Figures & Tables

Discussion

VS is a popular computational tool for identifying potential compounds against pre-determined biological targets [18]. The study focused on screening TCM compounds against the STAT3 protein. Initially, the selection of the top 20 compounds was based on their robust binding affinities with STAT3. Subsequently, SMARTS tools were utilized to filter this group, resulting in the identification of the top 6 compounds: ZINC85542844, ZINC4098720, ZINC85543599, ZINC85593523, ZINC85593528, and ZINC85593537. These compounds adhered to Lipinski and Veber’s criteria and display potential for future drug development. Moreover, further analysis was conducted on these compounds for an in-depth analysis of their interactions, revealing their binding capabilities with crucial residues of the STAT3 protein. The active site residues of STAT3 have been determined as Lys244, Asp369, Arg382, Arg423, Leu438, Gly442, Thr443, Glu444, Thr456 and His457 [19]. Consistent with this the hits (ZINC85542844, ZINC4098720, ZINC85543599, ZINC85593523, ZINC85593528, and ZINC85593537) were found to interact with these residues of STAT3 protein.

STAT3 protein residues, suggesting that these compounds bind within the same binding pocket of the STAT3 protein, similar to STX-0119.

Computational drug discovery is an effective strategy for speeding up and streamlining the drug discovery and development pipeline, demonstrating efficacy in increasing efficiency and cost-effectiveness [21-23]. In docking analysis, if a chemical has a lower binding affinity than the control, it indicates that the molecule is more active [24,25]. Interestingly, hits (ZINC85542844, ZINC4098720, ZINC85543599, ZINC85593523, ZINC85593528, and ZINC85593537) had higher negative binding affinity values relative to the control (STX-0119), indicating that these hits resulted in tighter binding and could be used as STAT3 inhibitors.

Natural products and structural counterparts have long played important roles in pharmacotherapy, particularly in the treatment of cancer and infectious disorders. With rising health risks, there is an urgent need to investigate preventive and treatment options, emphasizing the critical role of natural products in generating novel medicines and reducing disease spread and mortality. Plant-derived chemicals' therapeutic value has been recognized for decades, resulting in their widespread use in both households and clinics to treat a variety of illnesses [26].

This work used a structure-based drug discovery approach to identify natural compound-like inhibitors of STAT3. The hit compounds ZINC85542844, ZINC4098720, ZINC85543599, ZINC85593523, ZINC85593528, and ZINC85593537 demonstrated strong binding and stability with STAT3 and may serve as a lead for STAT3 inhibition to combat the cancers. However, further research is needed to optimize them as STAT3 inhibitors.

Conflict of Interest

The author declare that there is no conflict of interest.

References