In silico analysis of wild-type and mutant KRAS

The mutations of the KRAS gene at codons 12, 13, and 61 have been widely reported with different prognosis. In silico is one approach to explain the characteristics of the mutant genes. This study aimed to reveal the potential energy and fluctuations of the binding site and active site of wild-type KRAS (KRAS Wt) and mutant KRAS (KRAS Mt) at codons 12, 13


INTRODUCTION
Abnormalities in cell division are often associated with deviations of proteins involved in cell proliferation and differentiation, which are observable from any conformational changes in the geometric structure.
The Ras Protein and its family, including KRAS, NRAS, and HRAS, are proto-oncogenes, which play roles in transmitting surface receptor signals to effector protein through the PIK3/AKT and RAF/MEK/ERK pathways.These two pathways are responsible for the process of cell proliferation and differentiation.KRAS, NRAS, and HRAS can turn into oncogenes due to mutations in one or more of their base pairs that may increase cell proliferation and differentiation activities and, thereby, trigger resistance to epidermal growth factor receptor (EGFR) monoclonal antibodies (Knickelbein and Zhang, 2014).
Mutations at codon 12 have been reported in up to 80% of all occurrences, while mutations at codon 13 reach 17% and the rest occurs at other codons (Forbes et al., 2006).The KRAS gene mutations in pancreatic and colon cancer have been indicated at codon 12, changing the production of amino acid, that is, the substitution of cysteine for glycine (G12C) (Forbes et al., 2011).In patients with lung cancer, the mutations occur at the same codon as pancreatic and colon cancer, but the alterations in amino acid production involve the conversion of glycine to aspartate (G12D) (Stephen et al., 2014).Also, mutations have been identified at codon 13, i.e., where glycine changes to aspartate, and some other amino acids, and at codon 61 that influences the conversion of glycine to histidine.These mutations often take place in the GDP/GTP bond or called the GTPase domain.
The hydrolysis process can be described through the interactions between KRAS Wt and GTPase-activating protein (GAP) complex, specifically between the amino acids Q61 in KRAS Wt, as well as δ and β GTP, with the amino acid R789 in GAP.The NH 2 group of the amino acid R789 in GAP interacts with the CO group of KRAS Wt protein with a bond distance of 2.7 Å. The connection between the NH group of the amino acid R789 in GAP with O (δ and β) atoms in phosphate resulted from hydrolyzed GTP occurs with the bond distances of 2.8 and 3.2 (Å) (Scheffzek et al., 1997).Gao and Leif (2013) also illustrate that the side chain position of Y32 phenol of KRAS Wt protein shifts away from the δ-and β-phosphate of GTP, whereas R789 in GAP is closer to the δ-and β-phosphate of GTP.Y32 phenol group in KRAS Wt is within the distance of 7Å to the R789 in GAP, but it is shorter in KRAS Mt-D31N and D33N (±4 Å).Accordingly, Y32 in KRAS Wt is assumed to play an essential role in opening the gate between this group and the GTP binding site.
According to Futatsugi and Tsuda (2001), the position and orientation of Q61 affect the ability to activate the water molecule 175 (W175).W175, located close to the δ-phosphate in GTP, is considered as a requirement for the hydrolysis process of GTP to GDP.Krengel et al. (1990) also confirm this claim after observing KRAS Mt-G12R or G12V mutations.Aside from W175, Lysine 16

In silico analysis …(Frengki et al., )
91 has also been reported to be directly involved in the hydrolysis process of the phosphate groups of GTP to GDP when the distances between N atoms (Lys 16) and O1 atoms in the δ-phosphate and βphosphate of GTP are approximately 3 and 2 Å, respectively (Krengel et al., 1990).Mutations at codons 12, 13, and 61 are believed to cause changes in the orientation and distance of some of these amino acids.
The intrinsic GTPase activity of KRAS and its sensitivity to GAP can also be predicted from other parameters, such as the fluctuations of residues in the GTPase and the two binding motifs in KRAS, namely Switch I and Switch II.Potential energy can also be used in completing the prediction of GTPase activity.The characteristics of these two parameters are useful in assessing the prognosis and healing expectancy of cancer therapy.

MATERIALS AND METHOD Materials
In this experiment, some programs were used to analyze the mutation of codons.PyMOL v. 0.99 was applied for visualization, MOE 2007.09 for docking and homology modeling, and online CABS flex server for fast simulation of protein structure fluctuations.The data of the sequence of amino acids in KRAS enzyme (Homo Sapiens) was obtained from uniprot.org database (P01116).The KRAS protein template was downloaded from pdb.org (GDP ID: 1WQ1).

Methods
The sequences were transformed according to the mutated amino acids at codons 11, 12, 13, and 61 in FASTA format.The binding sites were identified using MOE, followed by creating KRAS Wt and KRAS Mt enzyme models (A11P, G12A, G12C, G12D, G13A, G13C, G13D, and Q61H).All models were evaluated structurally based on the root-mean-square deviation (RMSD) scores and plots of amino acid residues (Ramachandran Plots) (Petsko and Ringe, 2004).Both ligands and enzymes were prepared and optimized for their three-dimensional structures by adding hydrogen, removing water molecules, adding partial charges, and minimizing energy.Following this procedure was the docking process through the predetermined binding sites.The resulting enzyme-ligand complex was saved in .pdbformat, while the docking value was in .mdb.All homology modeling and docking processes were carried out in MOE software.

Data Analysis
The docking results were evaluated by visualization in MOE and PyMol software.Afterward, the analysis continued to the simulation of molecular dynamics for all KRAS before (pre-test) and after (post-test) the formation of the ligand-receptor complex.In this step, MOE was employed to observe potential energy changes, while the CABS-dock web server was used to identify the fluctuations of the constituent amino acids of KRAS Wt/Mt.These data were analyzed statistically.

RESULTS AND DISCUSSION
This research aimed to investigate the potential energy and fluctuations of the binding site and active site of KRAS Wt and KRAS Mt.The results showed that all types of modeled KRAS Mt had 99% structural similarity to the KRAS Wt gene with RMSD<1 Å.Based on the Ramachandran plot analysis presented in Figure 1, all models are considered to have good quality.

93
of non-glycine residues in the disallowed region (Petsko and Ringe, 2004).Therefore, all models were subjected to the subsequent analysis.
The KRAS models in Figure 1 show similarities to each other, and the difference mainly lies in the sensitive areas, such as the GTP binding site (codons 11-16), the switch I region (codons 30-38), and the switch II region (codons 60-68).The atomic fluctuations of the GTP binding sites in KRAS Mt and KRAS Wt tended to decrease, as listed in Table I.The decreased residual fluctuations in the GTP binding sites creates a more stable affinity with the ligands.According to Futatsugi and Tsuda (2001), the Lys 16 residue plays a crucial role in forming the KRAS-GTP complex.Table I shows that the fluctuations of the domains of the GTP binding sites in every KRAS Mt model have a downward trend.On the contrary, the changes in the Switch I region of KRAS Mt are more significant than KRAS Wt.The fluctuations of amino acid residues in Switch I have a substantial role in the affinity for and the interaction with GAP (Tyr 32).The data is provided in Table II.Table II presents significant fluctuations of Tyr 32 in all KRAS Mt.These fluctuations are assumed to reduce the open conformation of Tyr 32, which allows the insertion of Arginine 789 (amino acid; GAP) and its interactions with δ and β-phosphate (GTP) (Gao and Leif, 2013).
Table III shows that the fluctuations of all amino acid residues in Switch II of KRAS Mt are more significant than KRAS Wt, except for 61 amino acid residues.Variations in these 61 residues in all models (KRAS Wt and Mt) are similar, except for KRAS Mt-G12A and Q61H.Q61H fluctuates almost two times higher than KRAS Wt, which lowers its affinity for and its stable interaction with GAP (Scheffzek et al., 1997).It also reduces the ability of KRAS Mt-Q61H to activate the water molecule 175 that plays a major role in the hydrolysis process of δ-phosphate (GTP) (Krengel et al., 1990).The low interaction between KRAS Mt-Q61H and GAP decreases the catalytic activity of GAP in hydrolyzing GTP (1.9%) (Hunter et al., 2015).

Table III. The fluctuations in the Switch II regions of KRAS Wt and Mt
The differences in KRAS structures also affect the variations of the ligand-receptor affinity, as observed from the docking scores and the molecular dynamics simulations.The higher the docking score (more negative), the stronger the affinity for GTP (Hardono et al., 2013).Referring to the docking score, KRAS Wt has a relatively weaker affinity compared with KRAS Mt, except for G13C mutant.The ligand-receptor complex formed between GTP and KRAS Mt is stronger, and the signaling function of the messenger is expected to be more active than KRAS Wt.An increase in signaling function leads to uncontrolled cell growth and proliferation.KRAS gene mutation has been reported in almost 30% of cancer cases (Karnoub and Weinberg, 2008).It has also been detected in 70-80% of pancreatic cancers, 40% of colon cancers, and up to 30% of lung cancers (Stephen et al., 2014;Garcea et al., 2005;De Roock et al., 2010;Forbes et al., 2009).KRAS gene mutation is associated with improved signaling function of the messenger.

Figure 1 .
Figure 1.The fluctuations of the binding site and active site and the Ramachandran plots of the KRAS Wt/Mt modelsAll models in this experiment meet the requirements of a good-quality structure, namely RMSD is <1 Å and the Ramachandran plot shows >85% of amino acid residues in the core region and <15%

Figure 2 .
Figure 2. The results of the KRAS Wt and Mt complex with the docking score of GT