In silico study of fucoxanthin as a tumor cytotoxic agent.

BACKGROUND: Fucoxanthin is a potential tumor cytotoxic compound. However, mechanisms underlying the activities are unclear. AIM: This in silico study aimed to predict the main mechanism of fucoxanthin; whether with its binding to p53 gene, CDK2, or tubulin.
MATERIALS AND METHODS: In silico was studied by using Autodock-Vina's algorithms. The mechanisms being analyzed by comparison of fucoxanthin and native ligands binding energies in p53 gene (1RV1), CDK2 (1AQ1), and three binding sites of tubulin (1JFF-paclitaxel, 1SA0-colchicine, and 1Z2B-vinblastine site).
RESULTS: Autodock-Vina's algorithms were valid, as re-docking the native ligands to their receptors showed a RSMD value less than 2 A with binding energies of -11.5 (1RV1), -14.4 (1AQ1), -15.4 (1JFF), -9.2 (1SA0), and -9.7 (1Z2B) kcal/mol. Docking of fucoxanthin to subjected receptors were -6.2 (1RV1), -9.3 (1AQ1), -8.1 (1JFF), -9.2 (1SA0), and -7.2 (1Z2B) kcal/mol. Virtual analysis of fucoxanthin and tubulin binding structure showed the carboxyl moiety in fucoxanthin make a hydrogen bound with 355Val (2.61 A) and 354Ala (2.79 A) at tubulin.
CONCLUSION: The results showed that binding energy of fucoxanthin could only reach the same level as with colchicine ligand in tubulin. Therefore, it may predict that the most probable fucoxanthin main mechanism is to bind tubulin, which causes microtubules depolimerization and cell cycle arrest.

Fucoxanthin is a carotenoid pigment that commonly found in marine algae. This compound has pharmaceutical properties, such as tumor cytotoxic agent.[1-4] The fucoxanthin cytotoxic activity is selective, as it found to be toxic to HeLa cells but nontoxic to human lymphocyte cells.[5] This selectivity action also found between leukemia and normal human cells.[6] Therefore, fucoxanthin is a prospective substance to be develop as a pharmaceutical antitumor agent. However, the mechanisms underlying fucoxanthin cytotoxic activities are unclear.[78]

Recent wet lab studies showed that there are several possibilities of fucoxanthin mechanisms action to inhibit the growth of tumor cell. In a study of fucoxanthin cytotoxic action to HeLa cells, the mechanism was considered by activation of p53 gene expression that caused cell apoptosis.[9] On another study, the mechanism was claimed by induction at cell cycle arrest with the compound binds to CDKs inhibitors through p21 gene.[10] Additionally, there is another possibility, which is a mechanism involves microtubules, a substance in cell that responsible for cell structure. Binding of an active compound to microtubules will disrupt its role and prevents the cell growth.[11-13]

Therefore, it may consider fucoxanthin has multiple mechanisms to prevent the growth of tumor cell. Several compounds, such as flavonoid Casticin, have a cytotoxic action with multiple mechanisms.[14] However, it is also probable that these multiple mechanisms are divided into a main mechanism and side effects. To identify this, in silico docking study is a powerful tool to do the analysis. Virtual mechanism screening has advantages over in vitro or in vivo analysis within cost and time with the same accurate results.[15] Besides being a useful computational tool in prescreening procedure to identify ligands that have similar biological properties, in silico docking studies can be used for the same ligand to different receptors, in order to identify its side effects.[16] Moreover, in silico studies will also reveal the binding between the targeted ligand to its receptor in three dimensional outlook, to raise understanding on the active site of bioactive compound.

Materials and Methods

Hardware and software

Docking calculations were carried out on branded HP PC running windows XP as the operating system, with Intel 3 GHz Core 2 and 1 GB memory hardware. The softwares used for docking preparation were Yasara ver 1.6 and Autodock ver 4.2. Meanwhile, Autodock Vina ver. 1.1 algorithms[17] were used for binding calculation. Virtual analysis of docking site was used by Autodock Tools ver 1.4 and ArgusLab ver 4.01.

Docking preparation

The subjected anticancer receptors in this study was p53, CDK2, and tubulin. The structures of those receptors were obtained in pdb files from the protein data bank (www.pdb.org). Each of receptor had been docked with native ligands, a known tumor cytotoxic agents. The receptors code for p53 gene and CDK2 were 1RV1[18] (with imidazole ligand) and 1AQ1[19] (with Staurosporine ligand). Meanwhile, the tubulin structure, which has three binding positions in its structure, as shown in Figure 1, was gained from three different pdb files. The pdb files were 1JFF[20] (with paclitaxel ligand), 1SA0[21] (with colchicine ligand) and 1Z2B[22] (with vinblastine ligand). Meanwhile, fucoxanthin structure file was gained from electronic Marinlit database.[23] In preparation step, hydrogenation and separation of native ligands for each receptor were carried out by using Yasara ver 1.6. Subsequently, Autodock ver 4.2 was used to charge the receptors and ligands and to identify the conformers of targeted ligand (fucoxanthin) and native ligands.

Figure 1

Three binding sites of tubulin: Colchicines site (a), vinblastine site (b), and paclitaxel site (c)

Algorithms validation and docking calculations

Docking of each ligand to the receptors was performed by Autodock Vina ver. 1.1 algorithms. The docking procedure used for rigid receptor and flexible ligand. Binding site in receptors was collected by Autodock Tools Ver. 1.4 automatic identification. The default settings were used for all other parameters. Algorithms validation was conducted by re-docking native ligands to their receptors. The algorithms are valid if the re-docking results have a root square mean deviation (RSMD) less than 2 Å (Angstrom) from original structure.[24] After that, the binding energies of fucoxanthin to each receptor were calculated and compared to those of native ligands. Visualization of the binding site after docking analysis was performed by Autodock Tools ver 1.4 and ArgusLab ver 4.01.

Results

Validation of Autodock-Vina algorithms found each of the native ligand conformer had a RMSD value within 1--2 Å to their beginning structures in receptors, as shown in Table 1. This means the Autodock-Vina algorithms used in this study were valid. Moreover, Table 1 also shows binding energies of native ligands and fucoxanthin to each receptor, which it revealed that binding energy of fucoxanthin could only be reached at the same level of colchicine in tubulin receptor. Virtual analysis of binding structure for fucoxanthin bound in tubulin at the colchicine site shown in Figure 2.

Table 1

Binding sites coordinates, RSMD value, and binding energy of native ligand and fucoxanthin to each receptor

Figure 2

Virtual modeling of fucoxanthin bound with tubulin in colchicine site (a) and hydrogen bound of fucoxanthin with receptor (b)

Discussion

Binding energy is correlated with the probability of affinity and stable bound between ligand and its receptor.[2526] Binding energy value may also predicts the bioactivity value for a ligand to the corresponding receptor.[27] Commonly, receptor gained from the protein data bank already docked with a native ligand, which has a specific binding energy value. If the binding energy value of subjected ligand lower than that of the native ligand, then it may predicts the subjected ligand is very active.[28] Moreover, if the value is the same, it may predicts the activity is in a level with the native ligands. But if it is higher than the binding energy of native ligand, it may consider that the binding between ligand and its receptor is more unstable and the bioactivity underlying the mechanism is lower. Therefore, comparison of binding energies between fucoxanthin and native ligands may predict the main mechanism of fucoxanthin as a tumor cytotoxic agent.

As the results showed that binding energy of fucoxanthin to p53 gene was higher than that of the native ligand (imidazole), it is probable the binding present between fucoxanthin and p53 gene is unstable. Although p53 gene activation is a common way to cause tumor cell apoptosis, there are several cytotoxic molecules that act independently to p53 gene activation.[29] The same result was obtained for CDK2 receptor. Therefore, it is also predicted that the probability of stable bound between CDK2 and fucoxanthin is also low. On the other hand, as the binding energy of fucoxanthin was found in the same level of colchicine to tubulin, fucoxanthin may has a stable bound with tubulin at the colchicine site. The binding structure analysis showed the oxygenated site of the cyclohexane ring in fucoxanthin played an important role in this binding mechanism.

Based on these results, this study predicted that the main mechanism underlying fucoxanthin cytotoxic activity is its action predominately as tubuline binding agents (TBAs) in colchicine site. Other wet lab analyses which showed an activation of p53 and CDK2 predicted have a lower affinity than fucoxanthin as TBA, as its binding value higher than that of the native ligand. A predominant mechanism in a multiple mechanisms pathway is very common. An example of this is curcumin, which can induce apoptosis or block cell cycle progression in a variety of cancer cell lines in several pathways, but predominantly via p53-dependent pathways.[30] The prediction of fucoxanthin as TBA is supported by flow-cytometric analysis which showed fucoxanthin can cause an arrest in the cell cycle.[31] Microtubule-depolymerizing agents, such as colchicine, are known for their ability to cause simultaneous G1 and G2 arrests.[32] Moreover, TBA in colchicine site will also inhibit the production of next pure tubulin and cause apoptosis of tumor cell.[3334]