Docking and molecular dynamic simulations of Mithramycin-A and Tolfenamic acid against Sp1 and survivin

Abstract

Therapeutic targeting of Sp1 transcription factor and survivin, are studied in various cancers due to their consistent overexpression. These markers result in poorer cancer prognoses and their downregulation has been investigated as an effective treatment approach. Mithramycin-A and Tolfenamic acid are two drugs with innate anti-cancer properties and are suggested to be able to target Sp1 through GC/GT DNA binding interference, however in-depth binding and mechanistic studies are lacking. Through docking analysis, we investigated Mithramycin-A and Tolfenamic acid in terms of their specific binding interactions with Sp1 and survivin. Through further molecular dynamics simulations including Root Mean Square (RMS) Fluctuation and RMS Deviation, rGYr, and H-bond analysis, we identified critical residues involved in drug interactions with each protein in question. We show Mithramycin-A as the superior binding candidate to each protein and found that it exhibited stronger binding with Sp1, and then survivin. Subsequent molecular dynamics simulations followed the same trend as initial binding energy calculations and showed crucial amino acids involved in each Mithramycin-A-protein complex. Our findings warrant further investigation into Mithramycin-A and its specific interaction with Sp1 and their downstream targets giving a better understanding of Mithramycin-A and its potential as an effective cancer treatment.

Graphical Abstract

1. Introduction

The major constraints in cancer therapy are cancer cells developing resistance to standard care. The baculoviral IAP repeat containing (BIRC) gene family proteins are known to cause multi-drug resistance and lead to poor survival in cancer patients [1–5]. While the efforts to find drugs that directly target BIRC family proteins have not been very successful, our laboratory is involved in exploring alternative approaches by exploring agents that can impact their transcriptional regulation. The objective of this study is to conduct in silico analyses employing two models (MD: .Molecular docking; MDS: Molecular Dynamics Simulations). MD and MDS were employed to understand the binding abilities of Specificity protein 1 (Sp1) and BIRC family protein, survivin (BIRC 5) and two agents that are known to inhibit Sp1 and survivin, Mithramycin-A (Mit-A) and Tolfenamic acid.

1.1. Sp proteins and Survivin:

Sp1 belongs to the Sp/Krüppel-like family and its overexpression is evident in a multitude of cancers and are often related with poor prognoses [6–10]. Sp1 binds with GC/GT rich sequences and with the exemption of their zinc finger motifs, which remain fairly conserved, have variable structures allowing for the interactions with an array of target genes. Implicated in survival, growth, invasion, and metastasis, Sp1 has wide spanning and complex effects in homeostatic and malignant cellular processes [11–13]. Sp1 variably binds to promoter and non-promoter regions, sometimes in competitive fashion, and are heavily modulated by posttranslational alterations, and its C-terminal multimerization-domain connected to activating multiple adjacent Sp-bindings [14, 15]. Sp1 binding sites tend to exhibit biphasic kinetics, rapidly reaching maximal binding or more slowly achieving maximal binding, at various promoter and enhancer regions. Factors including: regions that are enriched with enhancer/polycomb-repressed genes, high-quality and copy numbers of the Sp1 motif, and association of a number of co-binding factors and proteins involved in post-translational modification, result in complex Sp1 binding dynamics [15].

BIRC 5, also known as survivin, as an anti-apoptotic protein, downregulates caspase-3, −7, and −9 function through interactions with hepatitis B virus X protein and X-linked inhibitor of apoptosis. Governed partly by Sp transcriptional regulation, survivin has been correlated with aggressiveness and poorer clinical outcomes for many cancers and as such multiple therapies have been pursued in its targeting [16–18].

1.2. Tolfenamic acid; (TA) Mithramycin-A (Mit-A):

TA and Mit-A, a NSIAD and an antineoplastic antibiotic (respectively) both exhibit innate Sp targeting. The fact that Mit-A binds G-C specific DNA binding sequences preventing subsequent Sp binding has been well documented, however; how Mit-A decreases the expression of Sp1 and other potential interactions between Mit-A, Sp1, and survivin have yet to be shown in great detail [19–21]. TA has the ability to induce ubiquitin-mediated proteasomal degradation of various proteins and results in the degradation of Sp1 likely through the same mechanism, but direct binding and interactions between TA and Sp1 have also been scantly detailed [22, 23].

Therapeutic targeting of Sp1 and subsequent downstream targets requires drug candidates that can directly interfere with putative GC-rich binding sites such as TA, Mit-A, and several of its analogues [20, 24–26]. Deeper mechanistic studies into understanding Sp1 binding and their interactions with various targets will potentially lead to more precise therapeutic avenues in the variety of cancers where they exhibit upregulation. Herein, we detail the ability of Sp1 to bind survivin and the simulated binding of Mit-A and TA to each protein and the effect that drug introduction is exerting on the protein complexes.

While multiple studies already demonstrated the ability of TA and Mit-A to inhibit either Sp1 or survivin [3, 27–35], the current study is focussed on in silico analyses with MD and MDS to determine the binding abilities of Sp1 with survivin and TA and Mit-A with Sp1 and survivin. Similar studies were helpful to establish the sites of interactions[36]. The findings presented here are intriguing and hold practical value for enhancing our understanding of the mechanisms these two agents. Moreover, these results offer potential alternative or complementary approaches that could be harnessed to explore the interactions of these agents, as well as others in future, in the quest for developing safer and less toxic therapies.

2. Methodology:

2.1. Sequence-structure-functional (SSF) analysis

The SSF information of human Sp1, and inhibitor of apoptosis survivin was fetched from UniProtKB database with ID P08047 (SP1_HUMAN), and O15392 (BIRC5_HUMAN) respectively. The available experimental structures of Sp1 from Protein Data Bank (PDB) and other structural databases did not give satisfactory three-dimensional (3D) structure; hence the predicted structure was retrieved from AlphaFold protein structure database of EBI. The sequence length of Sp1 was 785 amino acids (aa). In order to perform various in-silico analyses, the domain region of Sp1 consisting of aa (619–785) was considered. The structure for survivin protein was obtained from X-Ray diffraction (PDB ID- 2QFA; resolution: 1.40Å), chain A from the PDB database [142 (5–141) aa]. BIOVIA Discovery Studio Visualizer was utilized for co-crystallization/macromolecules.

Binding site prediction

Binding regions of Sp1, and survivin were speculated using the Computed Atlas of Surface Topography of Proteins (CASTp)’s consensus outcomes, Grid-based Hemi pocket finder (GHECOM) and DEPTH tools and coordinates, such as x, y, and z, were set by using the AutoDock 4.2 software.

Fetching of drugs

The structural required information of TA (Compound ID: 610479) and Mit-A (Compound ID: 163659)were fetched from PubChem [37] in “Structure Data Format”. These structures were transferred to .pdb format, the preferred format for various docking tools.

Energy minimization

The energy minimization of 3D structure of Sp1 was carried out using Desmond module of Schrödinger Maestro v 2022.4 in order to get the compactness as the retrieved 3D structures were found to be scattered.

2.2. Molecular Docking

Followed published methods with some modifications[36]. The Glide (Grid-based Ligand Docking with Energetics), module of Schrödinger, LLC, New York, NY, 2022 is constituting of three modes 1. High Throughput Virtual Screening (HTVS) mode used for the screening of number large number of ligands. 2. Standard-precision (SP) mode used for screen ligands with unknown quality in very large numbers. 3. Extra-Precision (XP) mode used for docking of small number of ligand molecules in more accurately. For docking analyses of TA and Mit-A against all three targets in extra precision (XP) mode was employed. The ligand preparation was done using LigPrep followed by grid preparation using Receptor Grid Generation. Intermolecular hydrogen (H)-bonds, along with binding energy values, and other electrostatic and hydrophobic interactions were used to describe and process the best docked complexes for further computational investigation. Intermolecular connections between protein-drug complexes were demonstrated using Schrödinger’s ligand interactions module and LigPlot+tool (EML-LBI, Cambridgeshire, UK).

Simulations using molecular dynamics (MD) studies

MD serves as a powerful computational tool for measuring the motion of atoms/molecules with structure-function interactions (macromolecules) [38]. The system’s dynamic “evolution” is depicted by assessing interactive behaviour of the atoms and molecules for a given time [39]. To affirm drug binding modalities and offer a comprehensive view of the protein-drug interactions, we performed MD simulations of the highest-scoring drug-protein complexes using the Desmond programme, with a simulation time of 100 nanoseconds (ns). Processes such as minimization, heating, equilibration, and manufacturing were all part of the MD protocol. The MD procedure included the steps-minimization/heating/equilibration/ production run [40]. OPLS4 force field will be used to minimize protein-ligand complexes, and the resulting topology & atomic coordinates were measured mechanically [41]. SPC solvent model-orthorhombic box (15×15×10 Å) was used to completely enclose the compound/drug. After 0.15 M NaCl was added, the physiological pH returned to normal. We set up the water box with a PME-Particle Mesh Ewald boundary condition to maintain the distance of solute atoms above 10 Å to the boundary. Structural changes alongside dynamic behaviour of the protein (300K; 100 ns NPT ensemble simulation) were assessed through RMSD-Root Mean Square Deviation and RMSF-Root Mean Square Fluctuation plots. RMSD is used to evaluate how much a protein’s backbones have shifted from one structural configuration to another. Finding the protein or complex’s flexible area requires the RMSF technique [42]. The predicted (mode) ligand-binding at the protein-binding site is presented in the simulation interaction diagram illustration [43].

In vitro assays:

TA and Mit-A were procured from MilliporeSigma (St. Louis, MO). The antibodies of Sp1 and survivin we purchased from Santa Cruz Biotechnology (Santa Cruz, CA) and R & D Systems (Minneapolis, MN) respectively. The effect of TA or Mit-A on the expression of Sp1 and survivin was evaluated using Ewing sarcoma cells. TC71, CHLA9 and CHLA10 cells were obtained from Children’s Oncology Group (Research Resource Identifiers: CHLA9: CVCL_M150; CHLA10: RRID:CVCL_6583; TC71: CVCL_S882).and cultured as per the conditions recommended by the cell repository. Cells were treated with vehicle or optimized doses of drugs, TA (15 μg/ml ) or Mit-A (40 nM), previously determined by cell viability assays, for 48 h. Protein extracts were prepared and protein concentrations were determined by BCA assay. Subsequent Western blot analysis revealed expression levels of Sp1 and survivin[31].

3. Results

3.1. Annotation of binding sites and Grid values of targeted proteins:

The residues that were represented as a consensus of data from three different web servers are Lys 684, Cys 688, Glu 690, Cys 691, Pro 692, Lys 693, Arg 694, Phe 695, Met 696, Arg 697, Asp 699, His 700, Lys 703, His 704, Ile 705, Thr 707, His 708,711Lys, Val721, Gly722, Thr 723, Leu 724, Pro 725, Ser732, Glu733, Ser735, Ala 738, 739Thr,740Pro,741Ser, Leu743, Ile744, Thr745, Ala753, Ile754, Pro756, Glu757, Ile759, Ala760, Asn 764, Gln 771, Asn783, Gly 784, Phe 785 for Sp1.

Gln467,Ile625,Pro626,Gly627,Gly629,Trp644,His645,Gly647,Glu648,Arg649,Cys653,Trp655,Tyr657,Cys658,Gly659,Lys660,Arg661,Phe662,Thr663,Arg664,Asp666,Glu667,Leu668,Arg670,His671,Arg672,Thr674,His675,Met691,Arg692,Ser693,Asp694,Leu696,Ala697,Ile700,Lys701,Gln704,Asn705,Lys707,Gly708,His710,Ser712,Ser713,Thr714,Val715,Leu716,Leu735,Ile740,Leu769,Gln770,Leu771,Val772,Glu781 and the residues Arg 18, Glu40, Ile 74, Phe 86,Val89, Gln92, Phe93 (BIRC 5-binding site formation).

The grid analysis was performed using the popular docking software system AutoDock, which is used for screening the agents against targets of interest [44]. Allocation of Kollman charges for the protein was performed with ADT v.1.5. The details of the Sp1 and survivin grids are given below (dimensions, spaces, and parameters): the x-centering: −9.203, y-centering: 2.084, and z-centering: −10.011 for SP1; 45.571, 10.753 and 38.345 for survivin in order to promote the ligand’s or drug’s most open conformation.

3.2. Molecular docking

The binding energies and other interactive studies for TA and Mit-A against Sp1 and survivin proteins forming six different complexes is presented in Table 1 and Figures 1–2. The results of docking revealed the distinct binding energy of the drug-protein complexes. Among the varied conformations from the docking data and using GLIDE for the inter-molecular interaction, only the most stable configuration with the highest binding energy was during considered. The molecular docking analysis revealed that Mit-A had higher binding energy against both targets compared to TA. Both targets complexed with Mit-A were subjected to simulations studies to evaluate the binding modes of protein-drug interactions during specific periods of time.

Table 1: Molecular Docking Scores:

Molecular docking scores of Tolfenamic acid and Mithramycin against SP1, SP3 and Survivin.

Sl. No.	Target	PubChem CID	Drug	Binding Energy (kcal/mol)	No. of H -Bonds	H-Bond Forming Residues	Average Distance of H-Bonds (Å)
1.	SP1	610479	Tolfenamic acid	−4.099	2	Thr739	~2.444
2.	SP1	163659	Mithramycin	−10.498	11	Lys684, Lys693, Arg694, His708, Lys711, Glu733, Glu757, Pro692	~2.352
3.	SP3	610479	Tolfenamic acid	−5.121	2	Arg649, Leu771	~2.403
4.	SP3	163659	Mithramycin	−13.971	9	Lys660, Lys701, Lys707, Leu771, Gln704, Ser713, Glu667,	~2.061
5.	Survivin	610479	Tolfenamic acid	−1.617	3	Gln92, Glu40	~2.423
6.	Survivin	163659	Mithramycin	−8.199	12	Lys15, Lys78, Gln92, Glu36, Glu40, Val89, Lys91, Leu87, Glu94, Gln92	~2.259

Figure 1. Hydrogen bonding, electrostatic and hydrophobic interactions between Sp1 and Tolfenamic acid or Mithramycin A:

Intermolecular hydrogen bonding, electrostatic and hydrophobic interactions formed between (A) Sp1-TA complex and (B) Sp1-Mit-A complex. The images were obtained by ligand interactions module of Schrödinger.

Figure 2. Hydrogen bonding, electrostatic and hydrophobic interactions between Survivin and Tolfenamic acid or Mithramycin A:

Intermolecular hydrogen bonding, electrostatic and hydrophobic interactions formed between (A) Survivin-TA complex and (B) Survivin-Mit-A complex. The images were obtained by ligand interactions module of Schrödinger.

3.3. Trajectory analysis of MD simulations:

For this study, we performed MD simulations of Holo1: Sp1-Mit-A complex, and Holo3: Survivin-Mit-A complex, using the Desmond suite (Schrödinger Release 2022–4: Maestro, Schrödinger, LLC, New York, NY, 2022) to explore the kinetics, binding mechanism, and inhibitor specificity of each system. As mentioned previously, the structures of the target protein and protein-ligand complex were taken into account when generating the final docked structures. To assess the receptor structural rearrangements and the stability of the docked complexes with Mit-A, we employed 100 ns MD simulations. The systems’ (Apo, Holo) dynamic stability was measured using the RMSD profile (backbone atoms: 100 ns; Apo, Holo) (Figure 4).

Figure 4. Structural alterations and dynamic behavior using root mean square fluctuation:

Conformational stability of Apo and Holo states of Sp1, and Survivin protein throughout 100 nanoseconds (ns) time period of MDS. (A) Cα-RMSF profile of Sp1-Mit-A complex (B) Survivin–Mit-A complex.

The RMSD backbone graph of Holo1: Sp1-Mit-A complex revealed a stable trajectory with 70 ns simulation than its Apo state, Holo3: Survivin-Mit-A complex reflected mostly inconsistent deviations throughout the simulation time period. Holo1 state gave the RMSD value ranging from ~9.2 to ~13 Å from 70 to 85 ns, ~4.8 to ~7.2 Å from 70 to 85ns and ~4.4 Å to ~6.0 Å from 85ns to 100ns in the case of Holo2. The values for Holo3 ranged from ~4.2 to ~9.0 Å from 70 to 100 ns. This shows that Mit-A potentially inhibits both targeted proteins.

Using RMSF, we were able to confirm the RMSD result and track the residues’ variability, however the mobility of various residues was found in all states as per RMSF plots (Figure 4 -5C). Overall, the Holo2 state reflected higher fluctuations as compared to the Holo1 and the Holo3 state, which is possibly due to the interaction with Mit-A over the time of simulation. The residues between 15 to 25 in Holo1 and 45–55 in Holo3, displayed higher fluctuations in their Cα atoms compared to other locations which may be due to the interaction of Mit-A with Sp1 and survivin proteins. Approximately 10 C- and N-terminal residues showed larger variations across all states, but these can be safely ignored. Proteins having ligand-interacting residues are denoted by green vertical bars. This plot suggests that the binding of drugs-proteins could lower the mobility of residues in the Holo versus in the Apo state.

Figure 5. The Solvent Accessible Surface Area graph:

Conformational stability of Sp1 and Survivin protein throughout 100 nanoseconds (ns) time period of MD simulations. Radius of gyration (Rg) profile of (A) Sp1-Mit-A complex (B) Survivin-Mit-A complex.; Solvent accessible surface (SASA) analysis (C) Sp1-Mit-A complex (D) Survivin-Mit-A complex.

The overall compactness for all the states (Holo1-Holo3) and stability of Mit-A in the binding pocket of the desired three receptors during the 100 ns MD simulation was elucidated using properties such as radius of gyration (rGyr), as shown in Figure 5.

The radius of gyration (rGyr), which is equivalent to the principal moment of inertia, is used to quantify the “extendedness” of a ligand. The fluctuation graphs of rGyr vs. simulation time period duration show that after 60 ns, rGyr remains consistent over the simulation time in all states (Holo1, Holo2, and Holo3). Mit-A’s rGyr variation in the receptor binding pocket of SP1 ranged from ~8.5 to ~9.0Å and demonstrated the stable behavior of Mit-A over 60ns to 100ns MD simulation while Holo2 reflected rGyr values ~8.4 to ~9.2Å. Values of ~7.9 to ~8.2Å for rGyr indicated a more compact state for Holo3, suggesting that rGyr is inversely related to compactness and vice versa [45]. These outcomes of rGyr are well aligned by RMSF analysis.

Hydrophobic interactions facilitate the contact of amino acids with specific solvents. The occurrence of these interactions between the solvent and the protein’s core residues is directly proportional to the exposed surface area. The graph depicting SASA-Solvent Accessible Surface Area (Figure.5C; 5D) showed a reduction of accessible solvent surface (Holo states). Based on SASA’s findings, binding Mit-A can cause a shift in the hydrophilic-hydrophobic interaction areas, which could influence protein surface orientations by moving amino acid residues from the exposed to the buried region. From 60ns to 100ns MD simulation, the SASA graphs showed a majority of the Holo states presented a buried state. Holo1 represented SASA with ~400 to ~840 Ų, whereas for Holo3, the SASA value was ~600 Ų to ~620Ų. The SASA graphs of the Holo3 state showed lower values than Holo1 and Holo2 states during 60–100ns of MD simulations. Implying that the AA residues of Holo3 may change from accessible to the buried region. Binding of Mit-A to the protein surface can lead to an alteration of hydrophobic and hydrophilic interaction areas, potentially causing an orientation change in the protein surface.

3.4. H-bond analysis:

Schrödinger Version 2022–4 was used for plotting the intermolecular hydrogen bonds of Holo states during MD simulations (Supplementary data, Figure S5). Simulation of the Holo states showed variability of intermolecular H-bonds during the course of the simulation. Post-MD simulation analyses revealed three H-bonds in the case of Holo1, six in Holo2, and five in Holo3. The amount of H-bonds was directly proportional to the stability of drug-protein complex over the course of the simulation. The intermolecular hydrogen bonds of Holo1 were measured. The stacked bar chart of Holo1 in Supplementary data, Figure S5A shows that AA residues of Sp1 such as Lys703, and Ser732 may play a pivotal role in the binding and regulation of the protein. These residues are the most important AA residues for protein binding and activity. Residues of some proteins can engage in multiple interactions of the similar subtype with the ligand, resulting in histogram values exceeding 0.4. During the simulation of the Holo1 state, a consistent amount of intermolecular H-bonds was observed throughout the entire simulation period (Supplementary data, Figure S5B). Three H-bonds were represented in case of post MD of Holo1 (Supplementary data, Figure S5C). The stability of the drug-target complex was found to be correlated with the quantity of H-bonds observed during the simulation. Forming of H-bonds with simulations of Holo1 residues: Lys684, Lys693, Arg694, His708, Lys711, Glu733 and Pro692 were broken down, but novel H-bonds (Lys712, Gly722), hydrophobic interactions, and van der Waals interactions compensated (Supplementary data, Figure S5C). Apart from a few novel interactions, compensation was not observed at amino acid residue Glu757. This indicates that the efficacy of Mit-A against Sp1 could crucially depend upon Glu757 and could be a critical residue in inhibiting cancer progression.

Regarding survivin (Holo3), analysis of the stacked bar chart in Supplementary data, Figure S6A reveals that certain amino acid residues, namely Asp16, Val89, Lys90, and Gln92, likely have a significant impact on the binding and regulation of the protein. Additionally, it is worth noting that specific protein residues can engage in multiple interactions of the same subtype with the ligand, with values potentially exceeding 0.6 in the histogram. Furthermore, the simulation of the Holo3 state consistently exhibited a specific number of intermolecular H-bonds throughout the simulation, as depicted in Supplementary data, Figure S6B. During simulations of Holo3, H-bond forming residues: Lys15, Lys78, Glu36, Glu40, Lys91, Leu87, Glu94, Gln92 were broken down, but later novel H-bonds (Asp16, Lys90), hydrophobic interactions, and van der Waals interactions compensated (Supplementary data, Figure S6C). The residues Val 89 and Gln92 were not compensated for which indicates their importance in inhibiting cancer progression.

3.5. Expression of Sp1 and survivin

Consistent with published results on other cancer cells, both TA and Mit-A decreased the expression of both survivin and Sp1 in all three cell lines (Figure 6).

Figure 6. Effect of TA or Mithramycin-A on Sp1 and survivin expression:

Ewing sarcoma cells (TC71, CHLA9 and CHLA10) were treated with vehicle or TA (15 μg/ml ) or Mit-A (40 nM) and the expression of Sp1 and survivin was determined at 48 h post-treatment. Consistent with published results on other cancer cells, both TA and Mit-A decreased the expression of both survivin and Sp1 in all cell lines .C: Control; T: Tolfenamic acid; M: Mithramycin-A.

4. Discussion

Cancer therapeutics is an ever-changing field that is always in need of new targets to enhance the treatment repertoire for clinicians. Identification of protein targets can be expedited tremendously with the help of docking and molecular dynamics simulations[46]. In this study Mit-A and TA were evaluated in terms of their interactions with Sp1 and survivin. Both drugs interacted with proteins in question, however Mit-A consistently presented lower binding energy scores, indicating higher binding affinity than TA. Therefore, we further examined the relationship between Mit-A, Sp1, and survivin. Sp1 showed relatively strong binding interaction with Mit-A while survivin consistently showed slightly less efficient binding properties through all testing.

Among the Sp family members Sp1 and Sp3 both can regulate survivin and associated with cancer[19, 31]. In order to see if the interactions/affinities of other family members of Sp family, we further examined the binding-relationship with Sp3. Sp1 and Sp3 showed relatively similar strong binding interaction with Mitt-A with survivin consistently showing slightly less efficient binding properties through all testing. RMSD showed deviations in all structure complexes compared to apo states with eventual stabilization at different time points suggesting variations in the stability of the Mit-A-ligand complexes. These findings match initial binding energy scores with Sp3-Mit-A exhibited better stable interactions (Table 1; Supplementary Data Figures S1–S4). RMSF confirmed these findings with Sp3-Mit-A providing the highest degree of fluctuation at multiple residues. Sp1- and survivin-Mit-A complexes also showed fluctuation across various aa ranges indicating a decrease of residue mobility in all its Holo states. Continued confirmation of a higher degree of stability and efficiency of Sp1/Sp3-Mit-A complex in comparison to survivin was shown with more favorable rGyr and SASA simulations for the former. Decreased overall binding interactions in the survivin-Mit-A complex could be explained by lower SASA values.

H-bond analysis identified multiple crucial residues in each of the Mit-A protein target complexes. Interestingly, each complex showed evidence of residues that were uncompensated for post MD simulations indicating their importance in the ability of Mit-A to inhibit cancer progression. Glu757 (Sp1), and Val89 and Gln92 (survivin) all seem to play important roles in our complexes and Mit-A’s ability to inhibit cancer progression may hinge on these residues. While potential stable complexes are being formed between Mit-A and each protein in question, further confirmation of direct interaction will need to be tested in-vitro to confirm in-silico findings.

5. Conclusion

To date Mit-A has been known to be an effective Sp1 inhibitor,[19] however our objective is to offer a more in-depth explanation of Mit-A mechanistic abilities, lending more detail to potential effects on other protein targets. Our studies suggest that Mit-A exhibits better binding interactions with Sp1 and survivin compared to TA. Docking and molecular dynamics simulations suggest Mit-A interacts effectively with Sp1 and that could also extend separately to survivin. This study was focused on Docking and MD to gather initial evidence. Established approaches, such as molecular mechanics energies in conjunction with Poisson-Boltzmann or generalized Born and surface area continuum solvation (MM/PBSA and MM/GBSA), are commonly favored for assessing ligand binding with biological molecules. Consequently, for further validation of the binding capabilities of Mit-A and tolfenamic acid, future investigations should consider employing MM/PBSA and MM/GBSA methodologies.

Consistent with published data, Figure 6 shows the inhibitory effect of TA and Mit-A on Sp1 and survivin expression. It is plausible that these agents may cause a decrease in Sp1 expression through other mechanisms [27, 47], still it is critical to understand the transcriptional regulation of survivin. This could lead to the invention of alternative approaches for developing targeted agents for addressing tumor cells aggressive growth and the ability to develop resistance to standard care. Mit-A’s flexibility to interact efficiently with both proteins is an encouraging sign which also aligns with its anti-cancer activity in some aggressive tumor cells (e.g., Ewing sarcoma cells). While we only tested the ability of these agents to interact with Sp1 and survivin, similar models can be used to evaluate the interaction with other critical proteins associated with cancer and to further elucidate mechanisms of action.

Figure 3. Structural alterations and dynamic behavior using root mean square deviation:

Conformational stability of Apo and Holo states of Sp1, and Survivin protein throughout 100 nanoseconds (ns) time period of MDS. (A) Backbone-RMSD of Sp1-Mit-A complex; (B) Survivin–Mit-A complex.

Highlights:

Docking studies showed binding affinity for both Mithramycin and tolfenamic acid with Sp1 and BIRC5

Docking analysis confirmed that Mithramycin exhibits higher binding affinity in compared to Tolfenamic acid

Molecular Docking/Dynamic Simulation studies are consistent with in vitro results using Ewing sarcoma cells

Intermolecular H-bonds revealed novel bonds and interactions forming post Molecular Dynamic simulations

Data Availability:

Results were presented in manuscript and supplementary data. Any additional data will be available on request from the corresponding author.