INTRODUCTION:

Curcuminoids as the main secondary metabolites of temulawak and turmeric plants have traditionally been used as medicine in various countries such as India, China, and Southeast Asian countries including Indonesia. These compounds are reported to have broad biological activities such as antioxidant, neuroprotective, antitumor, anti-inflammatory, antiarthritis and as radioprotective.

Curcumin has even been used in first-stage clinical trials in the treatment of chronic diseases such as colon cancer, lung cancer, breast cancer and inflammatory-related diseases¹.

The molecular mechanism of curcumin as an anticancer includes many pathways including suppression of survival signals through inhibition of Akt. Curcumin also suppresses the inflammatory process through suppression of the inflammatory factor NFB which is followed by a protein cascade such as COX-2. Curcumin also triggers apoptotic events through intrinsic pathways in the form of cell stress that permanently damages DNA, devective cell cycle, or loss of growth factors that lead to the release of cytochrome C. While extrinsically curcumin has been reported to stimulate activation of Fas, tumor necrosis factor (TNF) and death receptors. (DF) which will form a complex with the Fas-associated death domain (FADD). This complex then activates caspase 8, 9, and finally caspase 3 as the executor of apoptotic events². Curcumin has also been reported to decrease VEGR angiogenesis factor T47D cells³, also in human intestinal microvascular endothelial cells through inhibition of COX-2 and MAPK pathways⁴.

The combination of Paclitaxel-curcumin has been reported by Kim et. al,⁵. The results of in vitro tests on pancreatic cancer cells showed that the anticancer effect of the combination of paxlitaxel and curcumin in a formula based on albumin nanoparticles was better and more stable. Furthermore, this combination has even reached the first stage of clinical trials. The results showed a superior effect of the paclitaxel-curcumin combination compared to paclitaxel-placebo⁶.

Interestingly, it turns out that curcumin metabolites such as dihydrocurcumin, tetrahydrocurcumin, hexahydrocurcumin and octahydrocurcumin which are the result of metabolism by the reductase enzyme have also been reported to have pharmacological effects similar to the above effects of curcumin⁷. Dihydrocurcumin in vitro is reported to show improvement in non-alcoholic fatty liver diseases (NAFLD) liver cancer cell models⁸. Tetrahydrocurcumin is reported to have antioxidant, anti-neurodegeneration, anti-aging, and anticancer effects, even Liu et al.,⁹ reported that the anticancer effect of tetrahyrocurcumin was stronger than that of curcumin in the H22 cancer cell model. The molecular mechanism is reported to trigger P53 activation as well as MDM2¹⁰ inhibition. Antitumor effects were also demonstrated by hexahydrocurcumin and octahydrocurcumin through activation of apoptotic factors and inhibition of the function of NFB transcription factors in vivo^7,11,12.

This article will prove the potential of curcumin metabolites as anti-cancer as evidenced by the strong interaction of curcumin and its metabolites through the in silico method against several receptors that are targets of curcumin, such as cell proliferation and survival receptors (Akt, MAPK), receptors that are transcription factors. (NFkB), proinflammatory receptors (COX-2), apoptosis-associated receptors (P53 and MDM2), and metastases-associated receptors (VEGF). Then proceed to analyze the effect of curcumin or its metabolites on the affinity of paclitaxel to the “human tubulin -1” receptor which is the target of paclitaxel's action. Paclitaxel was chosen because it is one of the first-line options for the use of chemotherapy in almost all types of cancer. In silico the use of combination chemotherapy of paclitaxel with curcumin/its metabolites can be observed through an increase in the affinity of post-combination chemotherapy on its target receptor. QSAR (Quantitative structure activity and relationship) methods, molecular homology and molecular docking were used to explain the hypothesis.

MATERIALS AND METHODS:

MATERIALS:

The device used is an Intel dual core 2.1 GHz hardware processor with a memory capacity of 2GB and a professional Windows XP operating system and is equipped with internet access. While the software used is ChemOffice 10 (Cambridge) and Molecular Operating Environment (MOE 2008 version. 10) (developed by Chemical Computing Group, Inc. (Montreal, Canada)).

The QSAR data refers to the report of Meng, et al.,¹³ covering 20 curcumin derivative structures "training set" and 4 curcumin derivative structures "test set" which have been tested for their anticancer activity in the prostate gland experimentally. Molecular docking data were obtained using ligands constructed from PubChem's “SMILES” structure using the “MOE Builder” tool. The 3D structure of the ligand is then saved in mdb format.

Likewise, the MAPK receptors (pdb id. 3GCU), NFkB (pdb id. 3GUT), Akt (pdb id. 4GV1), MDM2 (pdb id. 4XXB), COX-2 (pdb id. 3LN1), and VEGF (pdb id. 3HNG) downloaded from the website of the 3D structure provider www.rscb.org, while the 3D structure of “human tubulin-1” was modeled based on UniprotKB-Q9H4B7 sequence data using the “MOE Homology molecular” tool. Receptors are saved in pdb format.

METHODS:

QSAR:

The 3D structure of the curcumin derivative was created using “Chemoffice 10” software and saved in pdb format. The 11 descriptors were chosen to represent the hydrophilic, hydrophobic and steric properties of the curcumin derivative compounds. They are AM1_dipole, AM1_HF, AM1_HOMO, AM1_LUMO, logS, mr, ASA_H, vdw_vol, glob, and logP (o/w).

The QSAR equation model was validated using an external validation technique (4 curcumin derivative “test set” structures) and leave one out (LOO) cross validation (20 curcumin derivative “training set” structures) characterized by cross validation coefficient (Q2). A QSAR model is considered good if the value of cross validation (Q2) > 0.5¹⁴.

Molecular Modelling:

The structure of the 3D sequence “human tubulin-1” (UniprotKB-Q9H4B7) was designed based on the template (pdb id. 1SAO) obtained from the article Sinha et.al.,¹⁵ using the program “MOE Homology molecular” version 10. 2008. The UniprotKB-Q9H4B7 data sequence in the form of “FASTA” is first inputted into the MOE window. The same process is also carried out on the 3D template structure. The display of the protein sequence to be modeled and the template protein sequence will be displayed sequentially in the MOE "Sequence Editor" window. Next do the "Alignment" and finally do the modeling through the "Molecular Homology" technique. The model obtained is evaluated using the "Ramachandan Plot".

Molecular Docking:

1. Docking file preparation:

Docking file preparation is done by optimizing the geometry and minimizing the energy of the three-dimensional structure of the ligands and receptors using MOE software which is run on a single dual core computer.

Geometry optimization and receptor energy minimization:

The process of optimizing the geometry and minimizing the energy of the receptor is done with MOE software. The initial stage is the addition of hydrogen atoms and protonation using the 3D protonate function. After that, the partial charge setting is carried out using a partial charge, with the parameter method used is the current force field. Furthermore, energy minimization was carried out with the MMF94x force field, the solvation used was gas phase, and the RMS gradient was 0.001 kcal/mol using the default and the output file was in .moe format.

Geometry optimization and ligand energy minimization:

Ligand optimization was carried out on MOE database viewer (dv) with mdb format. The optimization process begins with washing the ligands. Then, the partial charge of the ligand was adjusted using partial charge, with the parameter method used was MMFF94x. Solvation used during the ligand optimization process is in the form of gas phase. After that, the energy minimization process was carried out with an RMS gradient of 0.001kcal/mol

2. Receptor-ligand docking:

The docking process begins with tracing the binding site and selecting the amino acid residues that are the docking target. Followed by opening the ligand candidate file in mdb format. After the preparation is complete, docking is done with the simulation-dock program. The placement method used is a triangle matcher with 1000 rounds. The scoring function used is London dG by displaying the best data as much as 30. Furthermore, from the 30 best data displays, refinement is carried out using a refinement force field with a population repetition size configuration of approx. 1000 corresponds to the MOE default. The display of the results of the entire docking process selected is the best 1 data. Other parameters conform to the default of MOE and the output file is in .mdb format.

3. receptor-ligand reDocking:

The process is the same as the “receptor-ligand docking” above, the difference is that the target receptor has previously complexed with curcumin or its metabolites.

4. Evaluation of docking results:

The bonding free energy of the docking results is seen in the output in mdb format. The selected ligand-enzyme complex is a complex that has a bond free energy value. Residual contact and hydrogen bonds that occur in the best docked ligand-enzyme complex were identified and analyzed in three-dimensional media using the MOE software ligPlot, then visualized in the ligand interaction program.

RESULTS AND DISCUSSION:

QSAR:

The QSAR equation obtaned is:

Exp data = 19.13301-0.39405 (AM1_dipole) + 0.01629 (AM1_HF)+1.04015 (AM1_HOMO) - 0.58652 (AM1_LUMO)+0.55058 (logS) -0.67804 (mr) - 0.01575 (ASA_H)+0.11450 (vdw_vol)+0.14756 (glob) -0.12192 (vol) +0.65632 (logP(o/w))

The above equation was obtained based on the relationship of physicochemical properties to the anticancer effect of experimental test results of curcumin derivatives on prostate cancer. The above equation was tested for validity through the leave one out (LOO) cross validation method based on the “training set” data above and obtained a Q2 value of 0.507, then tested with external validation using 4 other curcumin derivatives as a “test set” compound. The results of the external validation of anticancer activity obtained a nearly linear correlation between the anticancer effect of the experimental test results vs the anticancer effect of the predicted results of the QSAR equation calculation. The details are shown in table 2 below,

Table 2: External validation of anticancer activity of 4 curcumin derivatives "test set"

S. No	Ligands	AM1_dipole	AM1-HF	AM1_Homo	AM1_LUMO	LogS	Mr
1	38 pdb	6.4591	68.7484	-8.7005	-0.3545	-3.8826	11.3067
2	40 pdb	4.5167	111.4469	-8.9621	-1.0916	-4.0268	11.2151
3	47 pdb	7.5901	101.5261	-8.9417	-0.5408	-5.1563	12.4106
4	53 pdb	3.4877	51.3804	-9.5678	-0.2679	-2.2678	8.6701

Continue Table

ASA_H	Vdw_vol	Glob	Vol	logP(o/w)	Experiment data	Prediction data
0.8177	539.8851	0.0448	393.1250	2.6330	5.4690	5.6375
0.8078	519.6307	0.0689	377.7500	2.1390	6.0970	6.5638
0.8075	603.4155	0.1309	436.8750	3.7190	6.7210	7.1030
0.8430	410.5015	0.0716	296.2500	2.6020	6.2680	6.1169

Table 3: Free energy, hydrogen bond from docking ligands with MAPK, Akt, NFκβ, MDM2, COX-2 and VEGF

Compounds	Receptor
	MAPK		Akt
	Docking’score (kkal/mol)	Hydrogen Bond	Docking’score (kkal/mol)	Hydrogen Bond
Curcumin	-11.8663	Lys 53, Asp 168	-12.5358	Thr 291
Dihydrocurcumin	-13.8347	Ile 147(2), Arg 147, Gln 325	-11.8340	Lys 179, Lys 276, Thr312
Tetrahydrocurcumin	-12.1586	Arg 70, Lys 152	-11.8066	Thr291(3), Thr 195
Hexahydrocurcumin	-11.8894	-	-14.0166	Asp 292, Gly 294
Octahydrocurcumin	-13.1179	Hie 148, Lys 152(2), Asn 155	-11.9643	Lys 158(2), Glu 234(2), Phe 237
Native ligand	-12.0386	-	-11.8260	Glu 278, Lys 276, Leu 295
	NFκβ		MDM2
	Docking’score (kkal/mol)	Hydrogen Bond	Docking’score (kkal/mol)	Hydrogen Bond
Curcumin	-10.2274	Thr 450 (3)	-12.9355	Asp 294, Lys 19, Thr 73
Dihydrocurcumin	-10.1772	Hip 441, Lys 446	-11.8755	Tyr 302 (2), Glu 293
Tetrahydrocurcumin	-10.7491	Lys 445 (2)	-13.4582	Ser 51, Tyr 302
Hexahydrocurcumin	-11.4655	Lys 122(2), Val 358, Thr 443	-12.6820	-
Octahydrocurcumin	-12.2212	Thr 450 (3), Lys 446	-14.3256	Tyr302 (3)
Native ligand	-9.8647	Thr 443 (2)
	COX-2		VEGF
	Docking’score (kkal/mol)	Hydrogen Bond	Docking’score (kkal/mol)	Hydrogen Bond
Curcumin	-11.2521	Val 102, Hie 75, Tyr 341	-12.0901	Val 892, Phe 1041, Lys 861, Leu 1043
Dihydrocurcumin	-12.4139	Pro 500, Ser 516(3)	-11.2549	Lys 861
Tetrahydrocurcumin	-14.5559	Tyr 371(2), Arg 106, Ser 339, Tyr 341, Ser 516	-11.6455	Hie 1020, Asp 1040
Hexahydrocurcumin	-12.1882	Tyr 341, Tyr 371	-11.2666	Phe 1041, Lys 861 (2)
Octahydrocurcumin	-10.3450	Tyr 371 (2), Arg 106	-12.0432	-
Native ligand	-14.2871	Arg 106	-10.6094	Lys 861

ReDocking of paclitaxel to the curcumin ligand complex and its metabolites against the “human tubulin-1” receptor model:

Curcumin and its metabolites are also able to change the geometric conformation of the paclitaxel target receptor, especially in the “site binding” area, thereby increasing the strength of the interaction of paclitaxel on the “human tubulin -1” receptor, which is characterized by an increase in the docking score and an increase in the number of hydrogen bonds in the ligand-receptor complex formed as in table 4 and figure 3 below.

Table 4: Comparison of the affinity of the combination of paclitaxel with and without ligands (curcumin, dihydrocurcumin, tetrahydrocurcumin, hexahydrocurcumin and octahydrocurcumin) at the “human tubulin -1” receptor

Compounds	Docking’score Paclitaxel
	on “human tubulin β-1” receptor
	without	with
Curcumin	-8.5715 kkal/mol	-9.8538 kkal/mol
Dihydrocurcumin		-9.5072 kkal/mol
Tetrahydrocurcumin		-9.8186 kkal/mol
Hexahydrocurcumin		-9.4135 kkal/mol
Octahydrocurcumin		-9.6441 kkal/mol

Figure 3: Visualization of paclitaxel docking on the “human tubulin-1” receptor. Paclitaxel alone (A), paclitaxel in combination with curcumin (B), paclitaxel in combination with dihydrocurcumin (C), paclitaxel in combination with tetrahydrocurcumin (D), paclitaxel in combination with hexahydrocurcumin (E) and paclitaxel in combination with octahydrocurcumin (F).

DISCUSSION:

Prediction of anticancer activity:

Based on the Tropsha¹⁴ criteria, the value of Q2 = 0.507 with the correlation level (R = 0.976) has crossed the limits of the acceptable QSAR equation model. Likewise, the correlation value of the external validation results reached R = 0.9009, so that the obtained QSAR equation met the validity requirements in predicting the anticancer activity of curcumin and its metabolites. The prediction results of the QSAR equation above indicate that curcumin metabolites (tetrahydrocurcumin, hexahydrocurcumin and octahydrocurcumin) have stronger anticancer activity than curcumin with pIC₅₀ values of 4.1555, 5.7353 and 5.5926, respectively. Meanwhile, dihydrocurcumin is weaker than curcumin with a pIC₅₀ value of 6.6975. Curcumin itself has a pIC₅₀ value of 6.0850. These results prove that curcumin metabolites are also anticancer as curcumin, even with stronger potency. The in silico results are similar to the in vitro test results as reported by Pandey et al.,⁷.

Molecular Docking:

Curcumin and its metabolites (dihydrocurcumin, tetrahydrocurcumin, hexahydrocurcumin, and octahydrocurcumin) have a docking score similar to or even stronger than the native ligand (control) when forming a complex with the MAPK receptor. Stronger docking scores are indicated by an increase in the number of hydrogen bonds as in dihydrocurcumin which forms 8 hydrogen bonds and octahydrocurcumin which forms 4 hydrogen bonds at the MAPK receptor. The Akt receptor also showed that curcumin and its metabolites (except tetrahydrocurcumin) had higher docking scores than controls. However, the strength of the docking score is not positively correlated with the number of hydrogen bonds formed. Hexahydrocurcumin which only has 2 hydrogen bonds with Akt receptors through residues Asp 292 and Gly 294 has the highest score (-14.0166kcal/mol), on the other hand tetrahydrocurcumin with 4 hydrogen bonds but has the lowest docking score (-11.8066kcal/mol) (Table 3).

The NFκβ receptor showed superiority of curcumin and its metabolites compared to native ligands with the highest docking score (-12.2212 kcal/mol) achieved by octahydrocurcumin. The value of the docking ligand score for the NFkB receptor is strongly influenced by the number of hydrogen bonds formed. Octahydrocurcumin has 3 hydrogen bonds with Thr 450 and 1 hydrogen bond with Lys 446. Likewise, hexahydrocurcumin has a high docking score and also forms 2 hydrogen bonds through Lys 122 residues, 1 hydrogen bond each with Val 358 and Thr 443. Lower as in the native ligand and dihydrocurcumin each forming 2 hydrogen bonds (Table 3). The binding of the ligand to the Lys residue indicates the ability of the ligand to suppress the activation of NFkB as reported by Huang et al.,¹⁶ and Frengki et al.,¹⁷. They were docked a number of sesquiterpene lactone compounds, while Frengki (2019) carried out deoxyelephantopin and isodeoxyelephantopin compounds which are also derivatives of sesquiterpene lactones.

In the MDM2 receptor, the highest docking score was achieved by octahydrocurcumin with a value of -14.3256kcal/mol and the lowest docking score was obtained by the dihydrocurcumin ligand. These two ligands form 3 hydrogen bonds at their receptors. The tetrahydrocurcumin ligand also has a high docking score, but only forms 2 hydrogen bonds. Furthermore, the hexahydrocurcumin ligand also has a stronger docking score than dihydrocurcumin, even in the absence of hydrogen bonds. This shows that the docking score is not affected by the number of hydrogen bonds. This conclusion is reinforced by the report of Pantoro et al.,¹⁸ which states that the van de Waals binding determines the interaction of a number of curcumin derivatives with the MDM2 receptor.

The interaction of curcumin and its metabolites (except tetrahydrocurcumin) on the COX-2 receptor showed a relatively low docking score compared to the native ligand. The number of hydrogen bonds formed also does not determine the strength of the docking score. Hydrogen bonds are formed through the interaction of the ligand with residues Arg 106, Tyr 341, Tyr 371 and Ser 516(Table 3). The results of this docking are not much different from previous studies such as those conducted by Yuniarti, et al.,¹⁹. In his report, it is known that curcumin and its derivatives interact to form hydrogen bonds with COX-1 and COX-2 receptors through residues Arg 120, Tyr 355 and Tyr 385. When compared with the results that the researchers did, they only differed in the position of the residues (Arg and Tyr). This is reasonable because the researchers used the COX-2 receptor reference (pdb id. 3LN1) while Yuniarti, et al.,¹⁹ used the COX-2 receptor reference (pdb id. 3PGH).

At the VEGF receptor the number of hydrogen bonds also does not determine the strength of the docking score. However, at the VEGF receptor, it shows that curcumin and all its metabolites have superior docking scores compared to native ligands (Table 3).

Modeling of the “human tubulin -1” receptor:

Homology modeling is a reliable method for predicting the target structure and obtaining a three-dimensional structure that is homologous to the protein used (>40% homology)²⁰. From the modeling results with 1SAO protein as a template, it has 77.03% similarity with the sequence "human tubulin -1" to be modeled.

The quality of the model was analyzed using the Ramachandran Plot. In the Ramachandran Plot, clusters formed from several amino acid residues can show the form of the secondary structure of the protein formed. Each amino acid that makes up a protein will have an angle of phi (Φ) and psi (ᴪ), so that each amino acid residue can be described as a plot (coordinates). The quality of the protein structure can be known by looking at the plot of non-glycine residues located in the dihedral region that is prohibited (disallowed region). If the non-glycine residue in the disallowed region is more than 15%, it can be said that the protein has very poor structural quality (very unstable). The results of the analysis of the structure of the "human tubulin -1" sequence showed that 87.03% of the structure was included in the core region of the Ramachandran Map and there were 13 amino acid residues found in the disallowed region, but still within the limits of the requirements of a good quality model²¹.

Re Docking of paclitaxel to the curcumin ligand complex and its metabolites against the “human tubulin -1” receptor model:

Paclitaxel single treatment formed 1 hydrogen bond through the Glu 181 residue when forming a complex with the modeled “human tubulin -1” receptor with a docking value of -8.5715kcal/mol. The interaction of paclitaxel with the human tubulin -1 receptor that has bound curcumin or its metabolites is known as combination treatment paclitaxel. The affinity of paclitaxel in combination treatment with curcumin at the human tubulin -1 receptor increased to -9.8538kcal/mol. The curcumin combination paclitacel treatment also formed 1 hydrogen bond through Glu 181 as was the single treatment paclitaxel interaction, but the intensity of the curcumin combination paclitaxel hydrogen bonding strength was 74.9% higher than the single treatment paclitaxel 29.14%. Paclitaxel combination treatment with dihydrocurcumin also increased the affinity value to -9.5072 kcal/mol with the addition of 2 hydrogen bonds through the Ser 176 residue (H Acc & H Donor). Paclitaxel combination treatment with tetrahydrocurcumin increased the affinity value to -9.8186kcal/mol with the addition of 3 hydrogen bonds also through the Ser 176 residue (2H Acc & H Donor) with an intensity of <90%. Paclitaxel combination treatment with hexahydrocurcumin increased the affinity value to -9.4135 kcal/mol with the addition of 3 hydrogen bonds through residues Ser 176, Gln 15 and Asn 99. Likewise, paclitaxel combined treatment with octahydrocurcumin increased the affinity value to -9.6441 kcal/mol with the addition of 3 bonds. Hydrogen also passed through the Ser 176 residue (2H Acc & H Donor) with an intensity > 90% (Table

CONCLUTION:

Curcumin and its metabolites have the ability to interact quite strongly with a number of proteins that target anticancer receptors. Curcumin and its metabolites are also able to increase the strength of the interaction of paclitaxel with receptors. Although curcumin is metabolized in the liver, its metabolites still act as anticancer with equal or better strength as curcumin itself through in silico evidence using the QSAR and Molecular Docking methods.

AUTHORS’ CONTRIBUTIONS:

All the authors have contributed equally.

CONFLICTS OF INTEREST:

The authors declare that they have no conflicts of interest.

REFERENCES:

1. Amalraj A, Anitha P, Sreerag G. Biological activities of curcuminoids, other biomolecules from turmeric and their derivatives. A review article. J Tradit Complement Med. 2016 Jun 15; 7(2):205-233. https://doi.org/10.1016/j.jtcme.2016.05.005

2. Park W, A.R.M. Ruhul A, Zhuo GC, Dong MS. New Perspectives of Curcumin in Cancer Prevention. Cancer Prev Res (Phila). 2013 May; 6(5):387-400. doi: 10.1158/1940-6207.CAPR-12-0410. Epub 2013 Mar 6

3. Meiyanto E, Rosita M, Muhammad D. PGV-1 decreases angiogenic factor (VEGF and COX-2) expression on T47D cell induced by estrogen. Indonesian Journal of Pharmacy. 2006; 17(1): 1 – 6. http://dx.doi.org/10.14499/indonesianjpharm0iss0pp1-6

4. Binion DG, MF Otterson, P Rafiee. Curcumin inhibits VEGF-mediated angiogenesis in human intestinal microvascular endothelial cells through COX-2 and MAPK inhibition. Gut. 2008; 57:1509–1517. http://dx.doi.org/10.1136/gut.2008.152496

5. Kim B, Changkyu L, Eun SL, Beom SS, Yu SY. Paclitaxel and curcumin co-bound albumin nanoparticles having antitumor potential to pancreatic cancer. Asian Journal of Pharmaceutical Sciences. 2016; 11: 88-714. http://dx.doi.org/10.1016/j.ajps.2016.05.005

6. Saghatelyana T, Armen T, Naira J, Anna T , Hasmik P, Araxia H, Lidia H, Mikael A, Jurgen A, Andre-Robert R, Areg H, Alexander P. Efficacy and safety of curcumin in combination with paclitaxel in patients with advanced, metastatic breast cancer: A comparative, randomized, double-blind, placebo-controlled clinical trial. Phytomedicine. 2020; 70. 153218. https://doi.org/10.1016/j.phymed.2020.153218

7. Pandey A, Maya C, Shruti M, Pramod K, Pallavi S, Rupesh C. Reductive metabolites of curcumin and their therapeutic effects. Review Arcticle. Heliyon 6 (2020) e05469. https://doi.org/10.1016/j.heliyon.2020.e05469

8. Yu Q, Yayun L, Yufei W, Young C. Dihydrocurcumin ameliorates the lipid accumulation, oxidative stress and insulin resistance in oleic acid-induced L02 and HepG2 cells. Biomed Pharmacother. 2021; 103:1327–1336. https://doi.org/10.1016/j.biopha.2018.04.143

9. Liu W, Zhenbiao Z, Guosheng L, Dandon L, Hanbin C, Hongmei Y, Jiali L, Yuhong L, Jianhui X, Ziren S, Hongying C. Tetrahydrocurcumin is more effective than curcumin in inducing the apoptosis of H22 cells via regulation of a mitochondrial apoptosis pathway in ascites tumor-bearing mice. Food & function. 2017; 8(9): 3120–3129. https://doi.org/10.1039/C7FO00484B

10. Mohankumar K, Sankar P, Subhashree S, Vivek KS, Larance R, Akhil CB, Chelakkan SB, Mohane SC, Rukkumani R. Mechanism of apoptotic induction in human breast cancer cell, MCF-7, by an analog of curcumin in comparison with curcumin - An in vitro and in silico approach. Chemico-Biological Interaction. 2014; 210(5): 51-63. https://doi.org/10.1016/j.cbi.2013.12.006

11. Zhang ZB, Dan-Dan L, Jian-Hui X, Yan-Fang X, Zheng-Quan L, Yu-Hong L, Wei-Hai L, Jian-Nan C, Xiao-Ping L, Zhi-Xiu L, Zi-Ren S. Curcumin’s metabolites, tetrahydrocurcumin and octahydrocurcumin, possess superior antiinflammatory effects in vivo through suppression of TAK1-NF-κb pathway. Front. Pharmacol. 2018; 9 (OCT):1–12. https://doi.org/10.3389/fphar.2018.01181

12. Zhang ZB, Dn-Dan L, Jianhui X, Guesheng L, Jiangtau Z, Weihai L, Huilin L, Tiegang Y, Ziren S, Jianping C. Octahydrocurcumin, a final hydrogenated metabolite of curcumin, possesses superior anti-tumor activity through induction of cellular apoptosis. Food & function. 2018; 9 (4): 2005–2014. https://doi.org/10.1039/C7FO02048A

13. Meng X, Lianhua C, Fucheng S, Mingyuan L, Junjie J, Hongzong S, Yunbo D, Honglin Z. 3D-QSAR and Molecular Docking Studies on Design Anti Prostate Cancer Curcumin Analogues. Current Computer-Aided Drug Design 2000; 16(3): 245-256 doi: 10.2174/1573409914666181029123746

14. Tropsha, A. Best Practices for QSAR Model Development, Validation, and Exploitation. Molecular Informatics. 2010; 29(6-7), 476-488. https://doi.org/10.1002/minf.201000061

15. Sinha R, Ambarish V, Shankaracharya. A molecular docking study of anticancer drug paclitaxel and its analogues. Indian Journal of Biochemisty & Biophysics. 2011; 48(2): 101-105. doi: 10.1080/03772063.2021.1878063

16. Huang CC, Chiuping L, Chihyang C and Liefen S. Deoxyelephantopin a novel multifunctional agent, suppresses mammary tumor growth and lung metastasis and double survival time in mice. British Journal of Pharmacology. 2010; 159(19): 856-871. https://doi.org/10.1111/j.1476-5381.2009.00581.x

17. Frengki, Dedi PR, Fatma SR, Daan K, Henni V. Nfκβ Inhibition Mechanism of Deoxyelephantopin and Isodeoxyelephantopin with QSAR and Molecular Docking. IJPSR. 2019; Vol. 10(7): 3228-3233. http://dx.doi.org/10.13040/IJPSR.0975-8232.10(7).3228-33

18. Pantoro BI, Nancy MR, Siti K, Anna SV, Yuniati S. In Silico Analysis of Turmeric (Curcuma Longa) as a Murine Double Minute 2 Protein Inhibitor for Treatment of Glioblastoma Multiforme. Oceana Biomedicina Journal. 2020; 3(2): 145-158

19. Yuniarti N, Perdana AN, Aditya A, Sardjiman S, Zuliies I, Enade PI. In vitro and In Silico Studies on Curcumin and Its Analogues as Dual Inhibitors for cyclooxygenase-1 (COX-1) and cyclooxygenase-2 (COX-2). ITB J. Sci. 2012; 44(1): 51-66. https://doi.org/10.5614/itbj.sci.2012.44.1.5

20. Kalyaanamoorthy S, Yie-Ping PC. Structure-based Drug Design to Augment Hit Discovery. Drug Discovery Today. 2011; 16(17/18): 831-839. https://doi.org/10.1016/j.drudis.2011.07.006

21. Frengki, Dedi PR, Fatma SR, Daan K, Henni V. In silico analysis of wild-type and mutant KRAS. Pharmaciana. 2019; 9(1): 89-98. https://doi.org/10.12928/pharmaciana.v%vi%i.11384

Received on 06.07.2021 Modified on 14.10.2021

Research J. Pharm. and Tech 2023; 16(2):885-892.

DOI: 10.52711/0974-360X.2023.00150

Compounds	Pub Chem ID	“SMILES FORM”
Curcumin	2662	COC1=C(C=CC(=C1)C=CC(=O)CC(=O)C=CC2=CC(=C(C=C2)O)OC)O
Dihydrocurcumin	10429233	COC1=C(C=CC(=C1)CCC(=O)CC(=O)C=CC2=CC(=C(C=C2)O)OC)O
Tetrahydrocurcumin	124072	COC1=C(C=CC(=C1)CCC(=O)CC(=O)CCC2=CC(=C(C=C2)O)OC)O
Hexahydrocurcumin	5318039	COC1=C(C=CC(=C1)CCC(CC(=O)CCC2=CC(=C(C=C2)O)OC)O)O
Octahydrocurcumin	11068834	COC1=C(C=CC(=C1)CCC(CC(CCC2=CC(=C(C=C2)O)OC)O)O)O
Paclitaxel	36314	CC1=C2C(C(=O)C3(C(CC4C(C3C(C(C2(C)C)(CC1OC(=O)C(C(C5=CC=CC=C5)NC(=O)C6=CC=CC=C6)O)O)OC(=O)C7=CC=CC=C7)(CO4)OC(=O)C)O)C)OC(=O)C