Elucidating the therapeutic potential of indazole derivative bindarit against K-ras receptor: An in-silico analysis using molecular dynamics exploration

Abstract

Ras gene is frequently mutated in cancer. Among different subtypes of Ras gene, K-Ras mutation occurs in nearly 30 % of human cancers. K-Ras mutation, specifically K-Ras (G12D) mutation is prevalent in cancers like lung, colon and pancreatic cancer. During cancer occurrence, mutant Ras remain in activated form (GTP bound state) for cancer cell proliferation. In the quest for a potential K-Ras inhibitor, nitrogen-containing indazole derivatives can show promise as inhibitors, as they have numerous therapeutic properties like anti-inflammatory, anti-viral and anti-tumor. Furthermore, among various indazole derivatives, “Bindarit” is an important therapeutic compound which could have potential inhibitory action against K-Ras due to its structural resemblance with reference compound “Benzimidazole”. So, the current study is an attempt to find out the inhibitory effect of Bindarit against K-Ras activation by binding to a pocket which is adjacent to the switch I/II regions of the K-Ras receptor. AutoDock tool was used to investigate the binding affinity of protein ligand interaction and GROMACS package was utilised to assess their interactions in a dynamic setting. Bindarit shows better binding affinity than reference with binding energy of −7.3 kcal/mol. Upon ligand binding conformational changes take place, which could lead to the loss of GTPase activity. Consequently, further downstream signalling of the K-Ras pathway would be blocked and this could lead to the inhibition of K-Ras dependent cancer cell proliferation. However, further validation of present study can be done through experimental assay such as cytotoxic and protein expression analysis.

Graphical abstract

Highlights

• •This study will assist in developing a promising inhibitor for G12D K-Ras.
• •In-silico based therapeutic exploration of “Bindarit” against K-Ras receptor.
• •Inhibition of K-Ras dependent cancer proliferation by blocking GTPase activity.

1. Introduction

Ras, a human oncogene, is mutated in more than 20 % of human malignancies. RAS proteins act as molecular switches which cycle between a dormant GDP-bound form and an active GTP-bound state. Ras, in its GTP-bound state, interacts with a number of effectors after being activated through guanine nucleotide exchange factors (GEFs) [1,2]. On the other hand, mutant Ras has decreased GTPase activity, which keeps it in an activated conformation for a longer duration, hence boosting Ras-dependent signalling and cancer cell survival or proliferation [3,4]. Among different Ras isoforms, K-Ras is the most often mutated of the three Ras isoforms (K, N, and H) [4], which leads to the survival of various cancer types such as lung, colon or pancreatic cancer. The three hotspots (G13, G12 and Q61) exhibit notable differences in the mutation frequency, which leads to further distinguishing the Ras isoforms. Among all K-Ras variants, G12 mutations make up 83 %, G13 mutations (14 %), and Q61 mutations, which are uncommon (2 %) [5]. Additionally, there might be notable changes in the frequency of mutations within a single Ras isoform throughout different types of cancer.

In melanoma, G12 mutations are uncommon, while N-Ras Q61 mutations make up the most commonly mutated hotspot. In contrast, acute myeloid leukaemia favours N-Ras G12 mutations. G12 mutations predominate in K-Ras alterations in pancreatic ductal adenocarcinoma (PDAC), whereas Q61 and G13 mutations are less common in PDAC. However, colorectal adenocarcinoma (CRC) has a comparatively high prevalence of G13 mutations [6]. When compared to other G12 and G13 mutants, G12D has been demonstrated to possess an intermediate intrinsic as well as GAP-mediated GTP hydrolysis rate; in contrast, mutations like G12A considerably decrease intrinsic hydrolysis, while G12C displays wild-type levels [7]. The most common mutations among G12 mutated residues, mutates at cysteine (G12C, 14 %), valine (G12V, 23 %), and aspartate (G12D, 36 %) [8]. When glycine at position 12 is mutated to aspartate (G12D), a larger, negatively charged side group is projected, which hinders GTP hydrolysis, damages GTPase activity, and keeps K-Ras in its active (GTP-bound) form [9,10]. Moreover, K-Ras mutant which is mostly common in lung, colon and pancreatic cancer is most frequently mutated among other Ras isoforms and particularly with its G12D variants. Apart from most frequent occurrence of K-Ras G12D mutants subtype, it is also associated with very poor survival [[11], [12], [13]]. So, in the present study we are focusing on the K-Ras G12D mutation, as so far no success achieved for K-Ras G12D inhibition except MRTX1133 which is currently in clinical trial stage [14]. Such mutation results in the loss of GTPase activity, which prevents activated Ras-GTP from converting to Ras-GDP, and as a result, the Ras protein remains activated, which helps in cancer cell proliferation. Although for K-Ras G12C subtype inhibition, few of the recent trials has provided promising results [15].

In context of therapeutics, nitrogen-containing indazole derivatives which are found to have different pharmaceutical properties [[16], [17], [18], [19]] can be targeted against the K-Ras receptor as they have structural similarity with previously screened K-Ras inhibitor of imidazole derivative. Moreover, among several indazole derivative, Bindarit an important indazole-derived compound with IUPAC name (2-[(1-benzylindazol-3-yl) methoxy]-2-methylpropanoic acid) have several therapeutic properties such as anti-inflammatory and anti-tumor activity. Bindarit have therapeutic potential against acute pancreatitis [20], lupus erythematosus nephritis [21], coronary restenosis [22], prostate cancer [23] and breast cancer [24]. Furthermore, as indazole containing compound have structural similarity with previously reported K-Ras inhibitor, so similarly Bindarit has a structural resemblance with the reference compound Benzimidazole, a potential inhibitor of the G12D K-Ras mutant. So, in the present study Bindarit has been considered as a promising therapeutic compound to be targeted against the K-Ras receptor to prevent its activation which benefits cancer cell proliferation. An experimental or in-silico validation is needed to show the therapeutic potential of Bindarit against K-Ras receptor. So, in the current study using molecular dynamics analysis, therapeutic potential of Bindarit against K-Ras receptor has been explored.

2. Materials and methods

2.1. Protein and ligand structure preparation

The RCSB Protein Data Bank (http://www.pdb.org) has provided the 3D structure of the K-Ras protein (PDB ID: 4DSU) [25]. It was found to have one unique protein chain. In the 3D structure protein was bound with cofactor GDP and ligand Benzimidazole. The nonprotein heteroatoms (HETATM) were removed using PDB Goodies [26]. The missing residues GLU62, GLU63, TYR64, SER181, LYS182, THR183, LYS184, CYS185, VAL186, ILE187 and MET188 were added using MODELLER. After modeling, the addition of missing hydrogen and energy minimization was done using Swiss PDB-Viewer [27]. The 3D structure of ligand of choice “Bindarit” an indazole-based derivative was retrieved from pubchem database (Bindarit, PubChem CID: 71354). Furthermore, 3D structute of Benzimidazole was taken as a reference compound. Ligand and reference compound was energy minimized using discovery studio software (Biovia, D. S, Discovery Studio Visualizer. San Diego, 2021). Prepared protein and ligand were taken further for molecular docking.

2.2. Molecular docking

The Auto Dock Tools package with version 1.5.6 (The Scripps Research Institute, Florida, USA) produced the input files for the docking of the K-Ras receptor and the chosen ligand [28]. Except GDP which was treated as a component of the receptor, all hetatoms as well as water molecules were eliminated from the PDB file. All hydrogens were added for protein, and the Kollman-united charges method was used to calculate partial atomic charges. As an input file required for the docking procedure, the generated protein structure was saved in PDBQT format. Missing hydrogens which are associated with ligands were combined, and autodock tools calculated the Gasteiger charges. Following preparation, the files were archived in PDBQT format. Torsion angles of ligands were processed, and it was provided with spatial degrees of freedom, so that it could move around the receptor molecule. The ligand binding pocket was determined through PLIP server (https://plip-tool.biotec.tu-dresden.de/plip-web/plip/index) and PDBsum server (https://www.ebi.ac.uk/thornton-srv/databases/pdbsum/). Residues in binding pocket were ASP54, TYR71, THR74, LEU56, LYS5, LEU6, VAL7 and SER39. The locations of the ligands on the designated active sites as well as the free energies of protein inhibitors binding were determined in this experiment using docking simulation with AutoDock. Grid box dimension of 116 × 90 X 116 points was measured using AutoGrid module with grid spacing of 0.375 Å. To generate various ligand conformations, the Lamarckian genetic algorithm was used during docking. The conformer which was having the most favorable and stable interactions was chosen and further processed for molecular dynamics simulations.

2.3. Dynamic study of protein-ligand complex using molecular dynamics simulations

MD simulations were used to gain an enhanced understanding of the likely behaviour of final ligands within the Ras binding pocket. The GROMACS 2020.1 package was utilised to investigate ligand-protein interactions in a dynamic setting. The top docking pose of the compound was chosen and put within the binding pocket as the initial point of MD simulations. GROMOS96 54a7 force field was used for the parametrization of protein structure [29] and SPC (simple point charge) was the chosen water model. Protonation state was also maintained using GROMOS force field. GROMOS force field was selected because it works well with the GROMACS package and performs efficiently in capturing interactions between proteins and ligands and moreover it also works efficiently with well folded protein [30]. Automated Topology Builder (ATB) server (https://atb.uq.edu.au/index.py) generated the ligand topology file [31]. The structure file and topology file for the ligand of choice and reference compound were obtained from the ATB server and saved in the working folder. The protein-ligand complex was totally submerged in a dodecahedron box and solvated following the generation of protein and ligand topology [32,33]. GROMACS' solvation approach was used to populate this box with water model molecules. The complexes were then energy minimized using the steepest algorithm along with a tolerance of value 1000 kJ/mol/nm. Furthermore, the system passed through NVT and NPT conditions, respectively [34,35]. At 300 K and 1 bar, the temperature and pressure remain constant. Ultimately, the MDs were kept running for 100 ns and then trajectory analyses were done using the Xmgrace tool [36] for the exploration of root mean square deviation (RMSD), root mean square fluctuation (RMSF), radius of gyration (Rg), solvent-accessible surface area (SASA), hydrogen bonding, Principal component analysis (PCA) and Gibbs free energy change [37,38].

2.4. PCA analysis and gibbs free energy change

Principal components (PCs) are a reduced collection of uncorrelated variables that are created from the original set of correlated variables using principal component analysis (PCA), a dimension reduction technique. These PCs assist in identifying the crucial motions needed to comprehend the flexibility and functionalities of complicated, high-dimensional biomolecular systems like proteins. PC analysis determines the positional covariance matrix of the Cα atomic coordinates and its eigenvectors [39]. The gmx_covar module of GROMACS package was used to build the covariance matrix of the Cα atomic position. On the other hand if we look into free energy change, the conformational states that a protein can assume in the free energy surface during simulations are analyzed by the free energy landscape (FEL). Any modification to the system's thermodynamic characteristics causes the landscape to shift and the population of conformal states to be redistributed [40,41]. The first two major eigenvectors (PC1 and PC2), respectively, were projected from the MD trajectory data to compute the FELs. The gmx_sham module implemented in GROMACS generated the FEL models.

2.5. Screening of drug likeliness using ADME analysis

ADME analysis, an acronym for Absorption, Distribution, Metabolism, and Excretion, plays a pivotal role in the field of pharmacology and drug development. This comprehensive approach involves the study of how a drug is absorbed into the body, distributed to various tissues, metabolized by enzymes, and ultimately excreted. Understanding these fundamental pharmacokinetic processes is crucial for predicting the drug's bioavailability, efficacy, and potential toxicity. SWISS ADME tool which is freely available at www.swissadme.ch and developed by the Swiss Institute of Bioinformatics [42] was used for (adsorption, distribution, metabolism, and excretion) ADME analysis. Furthermore, pkCSM server (https://biosig.lab.uq.edu.au/pkcsm/) was also used for detailed analysis of ADME properties of ligands. Lipinski [43], Veber [44], Ghose [45], Muegge [46] and Egan [47] rules of five (RO5) screening were used to apply drug-likeliness candidature. By using total charge, TPSA, and violation of Lipinski's filter, the Abbot Bioavailability scores were calculated to predict the likelihood that a compound will have at least 10 % oral bioavailability. The drug distribution was characterized on the basis of permeability of blood brain barrier (BBB) and Cytochrome P450 model was applied to determine the metabolism of drug. Total clearance model and renal substrate were used for the prediction of excretion of the drug.

3. Results

3.1. Molecular docking analysis

To find out and analyze the binding affinity of the K-Ras receptor with indazole derivative Bindarit, docking analysis was done through the AutoDock tool. The least binding energy of the K-Ras receptor with Bindarit was found to be −7.3 kcal/mol and interacting amino acids were ARG73, GLN70, MET67, THR74, ASP54, LEU6, GLY75, LEU56, LYS5, VAL7, TYR71 and amino acids involved in hydrogen bond are ARG73, GLN70 and THR74. On the other hand, while docking reference compound (Benzimidazole) against K-Ras receptor, the least binding energy was found to be −4.7 kcal/mol and interacting amino acids were LYS5, LEU6, VAL7, SER39, ASP54, LEU56, TYR71, THR74 and amino acid involved in hydrogen bond is ASP54 as shown in Table 1 and Fig. 1 and binding orientation of ligand at receptor active site is shown in Fig. 2. Apart from hydrogen bonding, other interactions existed in protein-bindarit complex were van der Waals interactions involving GLY75 and LEU6 residues, Pi-Sulfur interaction involving MET67, Pi-Pi T-shaped interaction consisting of TYR71 and Pi-Alkyl interaction consisting of LEU56, LYS5 and VAL7. Figures were drawn using the tool discovery studio 2021. Binding orientation and common interacting amino acids among protein-reference complex and protein-ligand complex shows that both reference compound and ligand are binding in the same binding pocket of K-Ras receptor which is adjacent to the switch I/II regions of the K-Ras receptor. Upon ligand binding at K-Ras G12D binding site, conformational changes takes place which could lead to the loss of GTPase activity and further inhibition of K-Ras dependent downstream signalling. Docking analysis also suggests that Bindarit has better binding affinity than Benzimidazole.

Table 1.

Molecular docking analysis of Bindarit and Benzimidazole.

Ligands	2-dimensional structure	Binding Energy (kcal/mol)	Interacting amino acids	Hydrogen Bonds
Indazole derivative (Bindarit)		−7.3	ARG73, GLN70, MET67, THR74, ASP54, LEU6, GLY75, LEU56, LYS5, VAL7, TYR71	3
Reference compound (Benzimidazole)		−4.7	LYS5, LEU6, VAL7, SER39, ASP54, LEU56, TYR71, THR74	1

∗(Amino acids forming hydrogen bonds are highlighted in bold).

Fig. 1.

Interaction of K-Ras residues. (A) Bindarit (B) Benzimidazole.

Fig. 2.

Binding orientation of ligand and reference compound at K-Ras active site. (A) Bindarit (B) Benzimidazole.

3.2. ADME analysis

Drug likeliness among indazole derivative Bindarit and reference compound Benzimidazole was computed through ADME analysis using the SWISS ADME and pkCSM servers. Indazole derivative has no violations in Lipinski rule of 5 while reference compound has violations in Goose and Muegge rule. In the bioavailability score, the indazole derivative has a better score than the reference compound. In terms of blood brain barrier permeability, both reference and drug can cross the blood brain barrier. Furthermore, in the lead likeness parameter, Bindarit has no violations but the reference compound has 1 violation as shown in Table 2. So, by looking into different scores of ADME analysis in both the compounds, it is found that Bindarit has better drug likeliness than reference compound Benzimidazole.

Table 2.

Drug likeliness and pharmacokinetics properties.

Compounds	Reference compound (Benzimidazole)	Indazole derivative (Bindarit)
Lipinski #violations	0	0
Goose #violations	3	0
Veber #violations	0	0
Egan #violations	0	0
Muegge #violations	1	0
Bioavailability score	0.55	0.85
PAINS #alerts	0	0
Brenk #alerts	0	0
Leadlikeness #violations	No, 1, MW < 250	Yes
Synthetic accessibility	1	2.81
GI absorption	High	High
Blood brain barrier permeability	Yes	Yes
CYP1A2 inhibitor	Yes	No
CYP2C19 inhibitor	No	Yes
CYP2C9 inhibitor	No	No
CYP2D6 inhibitor	No	Yes
CYP3A4 inhibitor	No	No
Total Clearance	0.907 log ml/min/kg	0.086 log ml/min/kg
Renal OCT2 substrate	No	No

3.3. Molecular dynamics simulations trajectory analysis

MD simulations using GROMACS 2020.1 were done to comprehend the dynamics and structure of K-Ras upon ligand binding. To understand the conformational dynamics of K-Ras upon ligand binding, several MD trajectories including root mean square deviation (RMSD), radius of gyration (Rg), root mean square fluctuation (RMSF), hydrogen bonding and solvent-accessible surface area (SASA) and were analyzed from the 100 ns MD simulations using Grace.

3.3.1. RMS deviations (RMSD)

RMSD quantifies deviations present in the backbone atoms of proteins from the starting position till the end of the simulation. The RMSD results exhibited higher deviations in the protein without ligand and protein-reference-complex when compared with the complex of protein-ligand. Initially in apo protein there is a large deviation in the backbone dynamics till 40 ns, after 40 ns the proteins maintain a stable trajectory till the end of the simulation. While in protein-benzimidazole-complex there is a large deviation in the backbone dynamics till 50 ns, after 50 ns it maintains a stable trajectory till the end of the simulation. On the other hand in the protein-bindarit-complex, the trajectory adopts stable conformations with lesser deviations throughout the simulation as shown in Fig. 3. In the present study in comparison with apo protein and protein-reference-complex trajectory, the protein-ligand-complex trajectory shows better stability and throughout the simulation shows lesser deviations. Thus, an alteration in the RMS deviations from higher to lower suggests the conformational shift and better stability of K-Ras protein structures once binding with bindarit.

Fig. 3.

Analysis of RMSD plot.

3.3.2. RMS fluctuations

To determine how atomic deviations vary protein flexibility, the RMSF was calculated for ligand-free protein, protein-reference-complex and the complex of protein-ligand in 100 ns simulations. RMSF was investigated to understand that how binding of ligand affects the flexibility of protein residues during simulation. By analyzing the RMSF results, we found over the length of the protein, more residue fluctuations in the apo protein than protein-reference complex and protein-ligand complex. Furthermore, protein-ligand complex found to have lesser fluctuations than protein-reference complex throughout the simulation. The residue fluctuations were most significant in the C-terminus portion of the protein (residues between 170 and 190). While the RMSF profile for the corresponding residue decreases after the binding of the ligand as shown in Fig. 4. Thus, RMSF results confirmed more stability of the protein structures upon ligand binding.

Fig. 4.

Analysis of RMSF plot.

3.3.3. Radius of gyration (Rg)

To determine the compactness of protein structures in the presence of benzimidazole and bindarit, Rg was analyzed for both the protein-reference-complex and the complex of protein-ligand for 100 ns MD simulations. A complex's compact character is defined by its low radius of gyration (Rg) profile; on the other hand, a greater Rg profile indicates a flexible nature. Additionally, radius of gyration can be used to assess any changes to the structure and protein motif and Rg gives details on the distribution of total protein volume in a spherical state. In the present study, we found that the protein with reference compound has higher gyrations as compared to protein with ligand over protein's size. In protein-ligand-complex, initially, the trajectory exhibits higher gyrations till 30 ns. After 30 ns, the protein attains the equilibrium state and acquires more compact and stable conformations till it reaches the end of the simulation in the presence of a ligand as shown in Fig. 5. Therefore, the protein in the presence of the ligand is more compact and stable after 30 ns with lesser gyration than the protein in the presence of the reference compound till the end of simulation. Moreover, average backbone Rg values of protein-reference complex and protein-ligand complex were calculated to be 1.701 and 1.706 nm respectively with a P-value less than 0.05, which indicates that both the complexes are statistically significant as per the Kolmogorov-Smirnov test.

Fig. 5.

Analysis of Rg plot.

3.3.4. Hydrogen bonding analysis

Hydrogen bonds perform a very crucial role in determining a protein's structural stability during simulations. It also plays a significant part in keeping ligands and proteins together and help ligands in recognizing exact interactions with active sites. The maximum number of H-bonds was calculated to be 4. On average, the Protein-Bindarit complex maintained a continuous formation of 3–4 hydrogen bonds over different time scales of the simulation while the Protein-Benzimidazole complex maintained a continuous formation of 2–3 hydrogen bonds. So, Protein-Bindarit complex form more hydrogen bonds as compared to Protein-Benzimidazole complex over different time scales of the simulation. This infers the stability of protein conformation after binding of indazole derivative Bindarit as shown in Fig. 6.

Fig. 6.

Hydrogen bonding analysis. (A) Protein-Bindarit complex (B) Protein-Benzimidazole complex.

3.3.5. Solvent accessible surface area (SASA)

Solvent accessible surface area (SASA) measures the accessibility of proteins to surrounding water molecules and shows the interaction among complexes and solvents. The solvent effect is very essential in maintaining protein stability and it also act as a driving force for protein folding. Based on our analysis, we observed higher accessibility in the protein-reference-complex when compared to the protein-ligand-complex. Initially, both the trajectories maintain a similar pattern of SASA deviations, but after 30 ns the SASA value increases in the protein-reference-complex. On the other hand, the accessibility decreases for the protein-ligand-complex after 30 ns with a slight increment at the end of the simulation as shown in Fig. 7. Findings suggest that, in the protein-reference-complex, most of the residues present on the surface are highly exposed to nearby water molecules and this increases the SASA. On the other hand after the binding of Bindarit, there is a shift in the protein's conformations and this decreases the accessibility of the protein's surface residues. Moreover, most of the protein's active sites are engaged in binding with the ligand, therefore there is less exposure of K-Ras residues to the surrounding water molecules, so there is decrease in accessibility and increase in stability.

Fig. 7.

Analysis of SASA plot.

3.3.6. Protein-ligand contact analysis

The interactions between active site residues and ligands play a crucial role in ensuring the inhibitory efficacy of inhibitors [48,49]. To evaluate these interactions, molecular dynamics (MD) trajectories were analyzed to examine the protein-ligand interactions within the simulated complexes (Fig. 8). The Protein-Benzimidazole complex shows fewer hydrophobic interactions, limited hydrogen bonding, and weaker interactions profile compared to Protein-Bindarit complex. The H-bonding was limited (only at ASP54) with an interaction fraction of 0.22, while residues such as LYS5, LEU6, VAL7, SER39, LEU56, TYR71, and THR74 contributed to hydrophobic interactions within the active binding pocket with an average frequency of 0.28 (Fig. 8A). In comparison, the Protein-Bindarit complex exhibited strong hydrogen bonding by residues GLN70, ARG73, and THR74 at an average of 0.43. Hydrophobic interactions in this complex involved residue such as LYS5, LEU6, VAL7, ASP54, LEU56, MET67, TYR71, and GLY75 with an average score of 0.46 (Fig. 8B), suggesting their role in enhancing the stability of the complex during dynamic simulations. This stability is further supported by the number of atomic contacts (<0.6 nm) over time, where the Protein-Bindarit complex maintains consistent and stable contacts (Fig. 8C), while the Protein-Benzimidazole complex shows greater fluctuations reflecting reduced stability (Fig. 8D). Thus based on the contacts analysis, it can be inferred that the Protein-Bindarit complex demonstrates a stronger and more favorable binding profile compared to the reference compound.

Fig. 8.

Protein-ligand contact analysis (A) Residues interaction profile of Protein-Benzimidazole complex (B) Residues interaction profile of Protein-Bindarit complex (C) Atomic contacts in Protein-Benzimidazole complex (D) Atomic contacts in Protein-Bindarit complex.

3.3.7. Principal component analysis (PCA)

Principal component analysis (PCA) is used to resolve the large concerted motions during dynamics simulations. These motions are essential for understanding protein functions. We can visualize and understand the intricate dynamic behavior of biomolecular systems by concentrating on principal components. PCA assist in understanding the fundamental dynamics and facilitates the study of conformational transitions and protein folding. In the protein-reference complex, we measured large and expanded motions with increased covariance matrix values of 13 as shown in Fig. 9 A. However, in the protein-ligand complex, these motions were found to be well-clustered and more compact with a decreased covariance matrix value of 6 as shown in Fig. 9 B. A lower covariance matrix for protein-ligand complex suggests that with minimal fluctuations the system maintains its structure. Moreover, a lesser covariance matrix in PCA indicates a higher stability in the protein-ligand complex. Thus, PCA results demonstrate the thermodynamic stability of the K-Ras protein structures after ligand binding.

Fig. 9.

Principal component analysis. (A) Benzimidazole (B) Bindarit.

3.3.8. Gibbs free energy analysis

The PC analysis does not break down large-scale motions of proteins into discrete conformational states because its focus is on the motions of proteins included in an MD trajectory. So, in order to find out the conformational states sampled during simulations, analysis of free energy landscape profiles has been executed. It has been done using data consisting of the first two principal eigenvectors (PC1 and PC2) of the MD trajectory. The maximum or minimum conformational states that a protein can assume in the energy surface during the simulation's timescale are determined by free energy landscape. Conformations of the protein's free energy minima are depicted in blue, whereas meta-stable as well as intermediate forms are depicted in green and yellow. The colour red denotes a high-energy state of the protein. Here we observed more energy minima conformations in Bindarit bound form than Benzimidazole bound state as shown in Fig. 10. The higher number of energy minima conformations signifies the favorable binding of bindarit towards the K-Ras receptor [50].

Fig. 10.

Gibbs free energy analysis. (A) Benzimidazole (B) Bindarit.

4. Discussion and conclusion

Ras proteins are essential GTPase, which regulate cellular signalling pathways, and have emerged as potential therapeutic targets in the fight against cancer. Cell survival, differentiation, and proliferation are all significantly regulated by these GTPases. Unfortunately, mutant Ras variants are often detected in different types of cancer, resulting in unregulated cell division and tumour development. The complex nature of Ras protein has made attempts to target it for cancer treatment extremely difficult. Numerous therapeutic options outlined will have an impact on the responsiveness of each Ras variant due to its unique biochemical features. For instance, in one of the studies, RMC-4550 effectively inhibited SHP2 in cells harbouring K-Ras G12 mutations, and it was more efficacious against K-Ras G12C as compared to K-Ras G12V or K-Ras G12D [51]. Moreover, even among different species, the mutational spectrum of the K-Ras protein differs, and previously it has been shown for lung tumors that mutational patterns in human lung tumors did not correlate with mouse tumors. The biochemical distinctions between the different mutants and the mutational mechanisms that resulted in the mutations in vivo are probably the causes of these variations in K-Ras variants. For instance, K-Ras (G12C) is the most prevalent mutation among smokers, whereas K-Ras (G12D) is the most prevalent mutation among non-smokers [52].

Furthermore, while highlighting the influence of allele-specific changes on Ras subtypes, Haigis et al. showed that K-Ras (A146T) and K-Ras (G12D) display different biological and signalling outputs in the colon and pancreas, indicating that these tissue-specific phenotypes are caused by different RAS mutants engaging different signalling pathways and having allele-specific signalling properties [53]. New developments in drug discovery and research have shown viable approaches. Apart from direct inhibition of K-Ras researchers are looking into all possible ways to block K-Ras dependent cancer cell proliferation, such as targeting downstream regulators which includes raf, mek or erk proteins [54]. Moreover in recent studies, researchers are also targeting guanine nucleotide exchanging factor which is responsible for the activation of K-Ras protein [55]. Although for KRAS-G12C mutations, there has been a breakthrough with the growth of KRAS-G12C inhibitor Sotorasib [15,56] and Adagrasib [57,58] but for KRAS-G12D mutation, there has not been much success. So, the present study is an attempt to target the K-Ras receptor with G12D mutation which is the most prevalent mutant Ras subtype in cancers like pancreatic and colon cancer. Using an in-silico approach this study tried to target Bindarit an important indazole-based compound against K-Ras GTPase activation. In the current study, with a binding energy of −7.3 kcal/mol, it indicates that Bindarit has a higher binding affinity than the reference which has binding energy of −4.7 kcal/mol. Moreover, conformational changes occur upon ligand binding at K-Ras G12D binding site and which could result in the loss of GTPase activity and hence K-Ras dependent cancer cell proliferation would be blocked [2]. Furthermore, molecular dynamics simulations conducted in 100 ns shows that the protein structure can be stabilized by bindarit's interaction with K-Ras G12D receptor at their binding site. This GTPase inhibition will assist in the development of potential inhibitors against K-Ras-dependent cancer such as colon, pancreatic and lung cancer, although further validation will be required through experimental assays.

CRediT authorship contribution statement

Parmar Keshri Nandan: Writing – original draft, Methodology, Formal analysis, Conceptualization. Jayanthi Sivaraman: Supervision, Investigation.

Ethics approval

Not applicable: no human or animal subjects are directly involved in this research.

Funding

No funding was required for this study.

Declaration of competing interest

I have nothing to declare.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data will be made available on request.