Classes of Mormordicoside with potent and selective tumor cell growth inhibitory activity: prediction of Pyruvate kinase muscle isozyme 2 (PKM 2) and Anti-apoptotic Myeloid leukemia 1 (MCL1) inhibitor through machine learning.

doi:10.21203/rs.3.rs-2069876/v1

Download PDF

Research Article

Classes of Mormordicoside with potent and selective tumor cell growth inhibitory activity: prediction of Pyruvate kinase muscle isozyme 2 (PKM 2) and Anti-apoptotic Myeloid leukemia 1 (MCL1) inhibitor through machine learning.

https://doi.org/10.21203/rs.3.rs-2069876/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

The difficulty in treating cancer resides in the ability to target abnormal proliferation while protecting normal proliferation, a feat that necessitates a thorough comprehension of both the normal and malignant mechanisms that promote cell growth and proliferation. Targeting cell death signaling pathways such as glycolytic and mitochondrial apoptosis is hallmark of many cancers the aim in which this research is ready to evaluate.

Methods

Atomistic molecular dynamics simulation of top hits after molecular docking and ADMET profiling of the ligands were performed for main protease-hit complexes.

Results

Docking scores of ligands used against PKM2 ranges from – 9.36 to – 12.1 kcal/mol, wherein, Mormordicoside-F2 had the highest score (-12.1kcal/mol) performing better than the FDA approved drug Benserazide(-7kcal/mol). Likewise, the scores ranged between – 8.51 and – 12.05kcal/mol for Anti-apoptotic Myeloid leukemia 1 (MCL-1), with Mormordicoside-F1 being the highest ranked compound performing better than the FDA approved drug Venetoclax(-8.6 kcal/mol). The RMSD plots obtained depicted stable trajectories with consistent and minor fluctuations implying that the protein (PKM2 and MCL1) backbone underwent minor structural perturbations. In addition, several significant peaks of increased fluctuations (RMSF) were also observed, indicating their increased interaction potential implying that the ligands could adapt effectively in the binding pocket of the protein. SASA analysis results shows that ligands used retained inside their shallow binding pocket. The phylogenetic tree obtained implies likelihood of reoccurring result of the Insilco profiling.

Conclusion

This research unveils that Mormordicoside F1 shows acceptable stability with Anti-apoptotic Myeloid leukemia 1 (MCL-1), likewise Mormordicoside F2 against PKM2. These hits may offer a more advantageous repurposing alternative.

Cancer

abnormal proliferation

PKM2

MCL1

FDA

Mormordicoside

Docking

molecular dynamics simulation

Cancer is a group of diseases characterize by abnormal cell growths that can invade or spread to other parts of the body. Cancer can also be defined by uncontrolled cell proliferation. Cancer is one of the leading causes of death worldwide and the number of cancer patients is steadily increasing. There are over 100 different types of cancer known to affect humans, each classified according to the type of cell it was originally affected in. It was estimated that by 2030, 26 million new cancer cases and 17 million cancer deaths are expected each year [1]. Cancer hallmarks include maintaining proliferative signaling, evading growth inhibitors, defending against cell death, achieving replicative immortality, inducing angiogenesis, and activating invasion and metastasis to support human tumorigenesis. These processes are regulated by complex genetic circuits, often involving growth factor signaling, and cancer-associated proliferation can be either independent of this circuit or of the same growth-promoting signals that drive normal proliferation. Dependent on aberrant activation many normal processes also involve rapid cell proliferation, such as embryonic development, the immune response to infection, wound repair, and cell turnover in tissues such as the intestine [2]. The latter is a process regulated by a complex genetic circuit that often involves growth factor signaling, whereas cancer-associated proliferation is independent of this circuit or aberrant in the same growth signals that drive normal growth. The challenge in cancer therapy is to be able to target abnormal growth while suppressing normal growth. This requires a thorough understanding of both the normal and malignant mechanisms that drive cell growth and proliferation.

Pyruvate kinase muscle isoenzyme 2 (PKM2) is found primarily in the cytosol and catalyzes the irreversible transphosphorylation between phosphoenolpyruvate (PEP) and ADP to release pyruvate as the final step in glycolysis. It functions as an evolutionarily conserved glycolytic enzyme that produces ATP. Glycolysis is the key process of glucose breakdown, where glucose is broken down into two pyruvates. PKM2 acts as a protein kinase, regulates gene transcription in the nucleus [3], it reprograms oxidative phosphorylation to aerobic glycolysis, and promotes intra mitochondrial tumor cell proliferation and migration [4]. In addition, PKM2 has been found in the extracellular fluid of patients with gastrointestinal, pancreatic adenocarcinoma, lung, ovarian, and renal cell carcinomas [5]. Pyruvate kinase can exist as a tetramer or dimer (Fig. 1). The tetrameric structure is the active form with high binding affinity for PEP, while the dimeric form has low binding affinity for PEP and low activity. PKM1 constitutively oligomerizes into tetramers under physiological conditions, whereas PKM2 can exist as dimers or tetramers depending on the appropriate regulator. PKM1 and PKM2 undergo differential allosteric regulation and covalent modification [6]. The glycolytic intermediate fructose-1, 6-bisphosphate (FBP) preferentially binds to PKM2 rather than PKM1, thus increasing the affinity of PKM2 for PEP [7]. Besides FBP, many other metabolites, amino acids and small molecules are involved in regulating PKM2 activity [8]. Binding of small-molecule PKM2 activators to PKM2 promotes tetramer formation, thereby constitutively activating PKM2 and suppressing tumorigenesis. Post-translational modifications of PKM2, such as phosphorylation, acetylation, or oxidation, promote low activity of dimeric PKM2 [9]. Therefore, the activity of pyruvate kinase can be regulated by altering its conformation. The constitutive activity of tetramers like PKM1 allows them to function as pyruvate kinase in the cytosol, promoting the glycolytic process and energy production, whereas the less active dimeric PKM2 is involved in the accumulation of glycolytic intermediates and Promotes subsequent biosynthesis in tumor cells [10]. Importantly, PKM2 dimers are imported into the cell nucleus and function as protein kinases. Evidence supports a possible role of PKM2 in tumorigenesis. As an embryonic isoform, PKM2 is reactivated in tumors and overexpressed in several cancer types [11]. Deletion of PKM2 in normal cells leads to expression of PKM1 and induces growth arrest by impairing nucleotide production and subsequent DNA synthesis. Switching from PKM1 to PKM2 has been demonstrated in various cancer types, and reverse isoform switching from PKM2 to PKM1 inhibits aerobic glycolysis and reduces tumorigenesis in a nude mouse xenograft model [6].

Anti-apoptotic myeloid leukemia 1 (MCL-1) is an anti-apoptotic member of the BCL-2 family. Critical to cytochrome c release from mitochondria and subsequent caspase activation is the complex interplay of both pro- and anti-apoptotic proteins of the Bcl-2 family. There are several pro- and anti-apoptotic family members that can physically bind to each other, and the relative proportions of pro- and anti-apoptotic family members ultimately determine cell fate in terms of survival or death [12]. When an organism is exposed to certain stimuli that disrupt this balance, cell proliferation can overtake apoptosis and even lead to tumorigenesis. The intrinsic (or mitochondrial) pathway of apoptosis and the extrinsic pathway of apoptosis are the two apoptotic pathways. The ability to dampen cell death signaling pathways, such as mitochondrial apoptosis, is a hallmark of many cancers, including B-cell non-Hodgkin's lymphoma (B-NHL). Activation of mitochondrial apoptosis is tightly regulated by members of the B-cell leukemia/lymphoma-2 (BCL2) family of proteins through homotypic and heterotypic interactions [13]. The BCL-2 family of proteins share one or four homologous BCL-2 domains (BH 1–4) and are divided into three subgroups (Fig. 2). These BCL-2 family proteins regulate mitochondrial membrane integrity through complex interactions. MCL-1 inhibits mitochondrial outer membrane (MOMP) permeabilization and release of cytochrome C from mitochondria. MCL-1 is needed for the survival of many cells in the nervous system, T/B lymphocytes, and cardiomyocytes [12]. The extrinsic pathway is facilitated through death receptor activation. This activates the initiator caspases 8 and 10, which can directly trigger downstream executive caspases, such as caspases 3 and 7, to fully commit to apoptosis. In addition, caspase-8 and caspase-10 activate Bid, which activates Bak and Bax to induce MOMP. MOMP is a link between extrinsic and intrinsic signaling pathways [14].

Overexpression of the MCL-1 protein, or amplification of the MCL-1 gene, protects cancer cells from apoptosis and reduces their sensitivity to commonly used anticancer agents. This has been shown to be a mechanism of resistance to multiple cancer therapies, including radiotherapy., chemotherapy, and the BH3 mimetic targeting BCL-2/BCLXL. Therefore, MCL-1 is a highly promising target for tumor therapy [12].

Today, despite considerable efforts, cancer is still an active killer worldwide. The most common and effective methods of cancer treatment are surgery, chemotherapy, and radiation therapy [16]. However, these treatments have many limitations and drawbacks. Most cancer patients are diagnosed too late for surgery due to misdiagnosis and other factors. Chemotherapy and radiation therapy have serious side effects and complications such as fatigue, pain, diarrhea, nausea, vomiting, and hair loss. In addition, chemotherapy and radiation therapy can gradually make cancer cells more resistant to treatment. Therefore, there is a constant need to develop new, effective and affordable anticancer agents. Medicinal plants have become a popular alternative to cancer treatment in many countries around the world. Momordicoside is one of several related cucurbitan triterpenoid glycosides that can be extracted from bitter melon vine (Mormordica charantia), an important grape for medicinal purposes. Also various traditional medicines to treat bronchitis, gonorrhea, rheumatism, blood diseases, cholera, anemia, ulcers, diarrhea, gout, colic, helminths, dysentery, liver and spleen diseases, cancer, diabetes, etc. The main constituents of this plant include proteins, triterpenes, alkaloids, steroids, inorganic compounds, lipids, and antidiabetic, anticancer and antitumor, antibacterial, anthelmintic, antimalarial, antiulcer and immunomodulatory.

Ligands preparation

For docking against pyruvate kinase muscle isozyme 2 (PKM 2) and Anti-apoptotic MCL1, the PubChem online library (https://pubchem.ncbi.nlm.nih.gov/) was used to download natural derived Product Mormordicoside F1 PubChem ID: 44445566, F2 PubChem ID: 44445567, G PubChem ID: 91895422, I PubChem ID: 71717036 and K PubChem ID: 57330180.

The food and drug administration (FDA)-approved drug library (Venetoclax zinc15 ID: ZINC150338755 for MCL1 and Benserazide with zinc15 ID: ZINC3830273 for PKM2 was downloaded from ZINC15 database. This is public repository accessible online containing information of compounds and their biological activities. The 3D structures were downloaded in 3D SDT format from the database and converted individually to PDBQT files using OpenBabel-3.1.0. (table 1).

Protein preparation for docking

The crystallographic structure 6NU1 and 6QFM (figure 3) of a minimal resolution used for this study were retrieved from Protein Data Bank (www.rscb.org), after which the water molecules of these proteins were removed using BIOVIA Discovery Studio version 21.1., then the addition of Hydrogen bonds and Geisteiger charges were added to generate the PDBQT files using autodocksuite 4.2.6.86 (mgltools v 1.5.7). Discovery Studio version 21.1 and CASTp23 online tools was used to predict the active site of each protein, the predicted site were confirm by comparing with binding site of the initial literature.

Docking strategy

The optimized 3D structures of ligands (Mormordicoside F1, F2, G, I, K Benserazide and Venetoclax) and proteins were thereafter docked using the autodocksuite 4.2.6.i86 (mgltools v 1.5.7). standard protocol was used in calculated the Lamarckian genetic algorithm [17]. The grid map for docking calculations was centered on the target proteins according to [18]. In order to model non-bonded polar and hydrophobic contacts in the inhibitor site of both 6NU1 and 6QFM, BIOVIA Discovery Studio version 21.1 was used. Further docking procedures was carried out using standard docking procedures. Prior to pose ranking for fitness within each protein's binding pocket for continuum and discontinuous bond interactions, the molecular docking was assessed based on the free binding energies of the ligands with the corresponding proteins. Post docking analysis, 3D and 2D images generations was carried out using pymol, ligplus and BIOVIA Discovery Studio version 21.1. The three-dimensional crystal structures of the target proteins are displayed in Figure 3.

Molecular dynamics simulation

Using complexes with the highest negative binding energy from out docking results, MD simulation was carried out using standard procedure. Succinctly, the simulations were performed over a 100 Ps period using GROningen MAchine for Chemical Simulations (GROMACS) version 5.1.4 [19] Charm-gui was used to solvate, generate a water box and further prepare them using charmm force field for gromacs simulation. Finally these solvated and equilibrated complexes (three protein–ligand complexes) were put to production phase to generate trajectories of root mean square fluctuation (RMSF), Coulomb energy (short-range) root mean square deviation (RMSD), solvent-accessible surface area (SASA), and LJ (short-range) by the in-built GROMACS scripts. Graphs were plotted using Xmgrace tool.

ADMET profiling

To establish the ADMET profile of four ligands (ligand with highest negative binding energy and FDA approved), the in silico prediction of their absorption, distribution, metabolism, excretion, and toxicity was made possible by using bioinformatics resources [22]. The drug-likeness properties of Mormordicoside F₁, F₂Benserazide and Venetoclax were calculated by using the ADMETLAB 2.0.

Phylogenetic analysis

The genetic kinship between the two target protein sequences under study was compared with sequences of their nearest neighbors. Six PKM2 protein sequences (4G1N, 5X1V, 5X1W, 6V74,6V75,6V76), 10 MCL1 protein sequences (606G, 5FDR, 5W8F, 4WMR, 6OQB, 6OQC, 6OQD, 6OQN, 6FS2 and 6QB6) were aligned in Alignment Explorer of MEGA4 software. The MUSCLE (UPGMA) method was used for trimming and to confirm the sequence alignment. The evolutionary distances and history were calculated using the Maximum Composite Likelihood and Neighbor-Joining techniques, respectively. The tree's robustness was evaluated using bootstrap analysis with 1000 replications. [20].

Molecular docking

The docking scores of phytochemicals Mormordicoside F1, F2, G, K and FDA approved drug Benserazide and Venetoclax docked against each of the against pyruvate kinase muscle isozyme 2 (PKM 2) and Anti-apoptotic MCL1 are presented in Table 2. Against PKM2 the metabolites had docking scores ranging from – 9.36 to – 12.1 kcal/mol, with Mormordicoside F2 having the highest score (-12.1 kcal/mol), performing better than the FDA approved drug Benserazide with docking score of -7 kcal/mol while against MCL1, the scores ranged between – 8.51 and – 12.05 kcal/mol, with Mormordicoside F1 being the highest ranked compound. When compared with the FDA approved drug F1 Performed better having higher binding negative energy than Venetoclax (figure 4).

The interaction analysis suggested that the complexes that had a good binding affinity were the ones that formed the highest number of hydrophobic interaction. Unfavorable donor-donor was observed in MCL1-FI interaction complex and PKM2 – Benserazide Interaction complex, A single pi-carbon interaction was observed in Venetoclax-Mcl1 complex was also observed to be formed with residue lysine (LYS A:242) (table 3 and 4)

Table 2: Docking scores, number of conventional hydrogen bond and amino acid interactions for PKM2 and MCL1 ligand complexes

TARGET	PKM2						MCL1
Compounds (Mormordicoside)	F1	Benserazide	F2	G	I	K	F1	Venetoclax	F2	G	I	K
Docking score (kcal/mol)	-9.86	-7	-12.1	-10.89	-11.68	-9.36	-12.05	-8.6	-9.76	-10.43	-9.11	-8.51
Conventional hydrogen bond	3	2	2	3	3	1	5	3	3	3	2	1

Table 3. Interaction Analysis for the 4 Phytochemicals and Target Proteins MCL1

	Hydrophobic interaction			Pi-cation/stacking interactions			Unfavorable donor-donor
Compound	Residue	AA	Dist	residue	AA	Dist	Residue	AA	Dist
F1	A:285	SER	4.08	A:289	PRO	4.9 & 5.11	-	-	-
	A:288	GLU	4.0 & 4.28	A:286	PHE	4.97 & 6.90	-	-	-
	A:292	GLU	4.97	A:243	VAL	6.17 & 5.10	-	-	-
	A:242	ASP	3.25	-	-	-	-	-	-
	A:	GLN		-	-	-	A:106	ARG	6.66
Benserazide	A:44	ASN	5.05	-	-	-	-	-	-
	A:43	ARG	4.26	-	-	-	-	-	-
F2	A:252	HIS	4.47	A:231	MET	6.55	-	-	-
	A:253	VAL	2.94	A:234	LYS	6.42	-	-	-
	A:245	SER	3.09	A:248	ARG	4.53	-	-	-
G	A:256	ASP	3.36	A:234	LYS	3.75	-	-	-
	A:253	VAL	4.54	A:238	LYS	5.01 & 5.00	-	-	-
	A:252	HIS	3.03	A:235	LEU	3.71 & 4.19	-	-	-
	A:249	VAL	2.37	-	-	-	-	-	-
I	A:245	SER	3.48	A:235	LEU	3.45	-	-	-
	A:249	VAL	6.49	A:238	LYS	4.37	-	-	-
	A:253	VAL	3.22	A:231	MET	6.47	-	-	-
	A:252	HIS	3.95				-	-	-
K	A:228	PHE	4.05	A:227	ALA	6.02	-	-	-
	A:262	GLY	4.46	A:224	HIS	5.66	-	-	-
	A:263	ARG	4.39	A:220	VAL	5.66	-	-	-
	A:266	THR	4.50	A:231	MET	6.11	-	-	-
	-	-	-	A:253	VAL	6.05	-	-	-
	-	-	-	A:249	VAL		-	-	-

KEY: AA=amino acid, Dist= distance

Table 4. Interaction Analysis for the 4 Phytochemicals and Target Proteins PKM2

	Hydrophobic interaction			Pi-cation/stacking interactions			Unfavorable donor			PI-carbon
compound	Residue	AA	Dist	Residue	AA	Dist	residue	AA	Dist	residue	Residue	AA	Dist
F1	Residue	AA	Dist	A:353	LEU	6.4	-	-	-	-	-	-	-
	A:29	HIS	4.49	A:394	LEU	4.85	-	-	-	-	-	-	-
	A:390	TYR	4.77	A:319	ARG	3.71	-	-	-	-	-	-	-
	A:388	ALA	4.32&3.45	A:401	LEU	4.74	-	-	-	-	-	-	-
Venetoclax	A:248	ARG	3.98	A:289	PRO	5.29&4.71				A:244	LYS	5.97
	A:300	ARG	6.46	A:243	VAL	6.06	-	-	-	-	-	-	-
	A:255	LYS	4.69	A:251	VAL	5.47	-	-	-	-	-	-	-
				A:289	PHE	4.36	-	-	-	-	-	-	-
F2	A:424	Cys	3.77	A:40	ILE	5.33				A:447	ARG	3.89
	A:425	SER	4.25	A:504	LYS	5.15&4.94	-	-	-	-	-	-	-
	A:16	GLN	5.56	-	-	-	-	-	-	-	-	-	-
	A:448	ALA	3.10	-	-	-	-	-	-	-	-	-	-
	A:498	LYS	6.04	-	-	-	-	-	-	-	-	-	-
G	A:178	ASP	3.65	A:175	LYS	3.173	-	-	-	-	-	-	-
	A:329	GLN	5.95	A:299	ILE	4.15	-	-	-	-	-	-	-
	A:303	ALA	3.99	A:302	PRO	4.77	-	-	-	-	-	-	-
	A:295	GLY	3.90	A:175	TYR	3.65	-	-	-	-	-	-	-
	A:301	ILE	4.91	-	-	-	-	-	-	-	-	-	-
	A:298	GLN	4.19	-	-	-	-	-	-	-	-	-	-
I	A:179	GLY	3.55	A:299	ILE	4.47	-	-	-	-	-	-	-
	A:175	TYR	3.19	A:306	VAL	5.26	-	-	-	-	-	-	-
	A:303	ALA	3.09	-	-	-	-	-	-	-	-	-	-
	A:178	ASP	4.71	-	-	-	-	-	-	-	-	-	-
	A:298	GLY	3.04	-	-	-	-	-	-	-	-	-	-
	A:347	ASP	2.79	-	-	-	-	-	-	-	-	-	-
K	A:354	ASP	3.61	A:26	PHE	5.79	A:354	ASP	5.19
	A:311	LYS	4.03	A:353	LEU	3.61	-	-	-	-	-	-	-
	-	-	-	A:394	LEU	4.04	-	-	-	-	-	-	-
	-	-	-	A:389	ILE	3.80	-	-	-	-	-	-	-
	-	-	-	A:398	LEU	6.02	-	-	-	-	-	-	-
	-	-	-	A:319	ARG	4.55	-	-	-	-	-	-	-
	-	-	-	A:401	LEU	5.50	-	-	-	-	-	-	-

KEY: AA=amino acid, Dist= distance

Molecular Dynamics Simulations

The goal of the MD simulation was to more precisely introduce ligand-DNA patterns in a setting similar to that found in nature. This procedure also served to evaluate the ligand binding patterns with various charges. The MD simulation can display more specifics regarding how DNA and ligands interact. For more detailed and accurate investigation, the trajectory stability and flexibility of the SA and SSA complexes were confirmed by means of different properties, including root-mean-square deviation (RMSD), root-mean-square fluctuation (RMSF), solvent-accessible surface area (SASA), LJ (short range) or coulomb (short-range) energies between protein and ligands (figure 5 to 9).

ADMET profiling

Evaluation of ADMET profiles of the ligands Through the admetLAB 2.0, the result of the in silico ADMET properties of all the ligands, including the probability are presented in table 5. it was observed that Mormordicoside F1 and F2 contain two H-bond donor (not more than 5), five H-bond acceptor. A logP value of 4.317, rotatable bonds less than 10, Topological Polar Surface Area (TPSA) off 117.84 Å2 and The low logP value of both ligand indicated good absorption and permeation with higher hydrophilicity. Additionally, it was found to be non-cancerous, HIA positive, BBB positive, non-AMES toxic and Caco-2 permeable as well as having negative hERG inhibition, negative results for aromatase binding, micronuclear, and negative results for biodegradation (Table 5).

TABLE 5: in silico ADMET properties of all the ligands, including the probability

		Metabolism
	CYP1A2-inh	CYP1A2-sub	CYP2C19-inh	CYP2C19-sub	CYP2C9-inh	CYP2C9-sub	CYP2D6-inh	CYP2D6-sub	CYP3A4-inh	CYP3A4-sub
Ben.	0.008	0.052	0.023	0.057	0.007	0.121	0.004	0.221	0.006	0.027
ven.	0.065	0.143	0.816	0.072	0.899	0.926	0.17	0.815	0.919	0.915
F1	0.001	0.5	0.006	0.919	0.011	0.067	0.005	0.542	0.309	0.428
F2	0.001	0.5	0.006	0.919	0.011	0.067	0.005	0.542	0.309	0.428

	Physiochemical			Excretion		Distribution			Absorption
	LogS	LogD	LogP	CL	T12	PPB	VDss	Fu	Pgp-inh	HIA
Ben.	-1.356	-0.858	-2.562	13.045	0.908	48.03%	1.171	57.37%	0	0.028
Ven.	-5.951	4.302	7.203	1.585	0.012	101.00%	1.041	0.38%	1	0.003
F1	-4.288	4.384	4.317	1.399	0.033	99.37%	1.424	1.96%	0.132	0.644
F2	-4.288	4.384	4.317	1.399	0.033	99.37%	1.424	1.96%	0.132	0.644

				TOXICITY
	hERG	H-HT	DILI	Ames	ROA	FDAMDD	SkinSen	Carcinogenicity	EC	EI
Ben.	0.015	0.142	0.097	0.567	0.066	0.013	0.91	0.431	0.004	0.016
Ven.	0.942	0.972	0.953	0.316	0.821	0.851	0.251	0.426	0.003	0.003
F1	0.02	0.185	0.013	0.047	0.342	0.571	0.012	0.031	0.003	0.004
F2	0.02	0.185	0.013	0.047	0.342	0.571	0.012	0.031	0.003	0.004
	Respiratory	BCF	IGC50	LC50	SR-p53	MW	Vol	Dense	nHA	nHD
Ben.	0.787	0.46	3.523	3.529	0.067	257.1	239.355	1.074	8	8
Ven.	0.485	0.905	4.979	3.458	0.596	867.32	848.47	1.022	14	3
F1	0.751	2.719	5.17	6.899	0.524	632.43	662.218	0.955	8	4
F2	0.751	2.719	5.17	6.899	0.524	632.43	662.218	0.955	8	4
	TPSA	nRot	nRing	MaxRing	nHet	fChar	nRig	Flex	nStereo	NonGenotoxic_Carcinogenicity
Ben.	148.07	6	1	6	8	0	7	0.857	2	1
Ven.	172.03	14	8	9	16	0	50	0.28	0	1
F1	117.84	8	6	14	8	0	30	0.267	14	0
F2	117.84	8	6	14	8	0	30	0.267	14	0
	LD50_oral	Genotoxic_Carcinogenicity_Mutagenicity	SureChEMBL	NonBiodegradable	Skin_Sensitization	Acute_Aquatic_Toxicity	Toxicophores	QED	LD50_oral	Genotoxic_Carcinogenicity_Mutagenicity
Ben.	0	1	2	1	10	1	5	0.137	0	1
Ven.	1	5	0	5	3	1	4	0.076	1	5
F1	0	0	0	1	1	2	0	0.283	0	0
F2	0	0	0	1	1	2	0	0.283	0	0

KEY: Ben.= Benserazide acid, Ven.= Venetoclax

Phylogenetic Relationship

Phylogenetic relationship between the test protein and other homologous sequences and am outgroup: (A) PKM2, (B) MCL1 is presented in figure 10.

Based on affinity characteristics, one of the key techniques for determining favorable protein-ligand interactions is known as molecular docking [21]. In the current investigation, we screened Mormordicoside classes, library of bioactive molecules of Mormordica charantia and docked them onto the active site of Pyruvate kinase muscle isozyme 2 (PKM 2) and Anti-apoptotic MCL1. Against PKM2 Mormordicoside F2 having the highest score (-12.1 kcal/mol), performing better than the FDA approved drug Ben.serazide with docking score of -7 kcal/mol while against MCL1, Mormordicoside F1 was the highest ranked compound. When compared with the FDA approved drug F1 Performed better having docking score lower than Venetoclax. A complex is considered to be strong when the number of hydrogen bonds is greater with a few hydrophobic interactions, Pi-cation/stacking interactions, Unfavorable donor-donor and PI-carbon interaction. We subjected all the phytochemicals from Mormordica charantia and standard drugs docked WITH PKM2 and MCL1 separately to check for the number of interactions that each compound formed. The interaction analysis suggested that the complexes that had a good binding affinity were the ones that formed the highest number of hydrophobic interaction. Unfavorable donor-donor was observed in MCL1-FI interaction complex and PKM2 – Ben.serazide Interaction complex, A single pi-carbon interaction was observed in Venetoclax-Mcl1 complex was also observed to be formed with residue lysine (LYS A:242).

The RMSD of the backbone coordinates from starting values during simulation time is an indicator for stability and conformational changes of a typical target in a molecular simulation. The RMSD values of Ben.serazide-PKM2 and F2-PKM2 complex backbone were plotted over the 100 ps of the MD simulation, as shown in. As observed, the trajectories of strands were relatively stable towards the end of simulation. However, after energy minimization and slow heating to reach equilibrium at room temperature, the RMSD of both systems reached equilibration and fluctuated around an average value after 100 ps MD simulation, which highlighted less instability of the both protein-ligand complexes. Here we noticed that the minimum and maximum RMSDs obtained shows that all the samples were of very low magnitude, in ascending order ranging from, 0.05 nm to 0.125 nm for Venetoclax-MCL1, 0.05 nm to 0.15 nm for F1-MCL1, 0.05 to 0.16 nm for Ben.serazide-PKM2, 0.05 nm to 0.168 for F2-PKM2. FDA approved drugs has the lowest fluctuation among all the protein-ligand complexes. A sharp increase was observed at the beginning for all the protein-ligand complexes. These low magnitude results are a strong indication of conformational stability that the protein had achieved with the ligand molecules. Similar to this, lower RMSF fluctuations imply less flexibility, which reduces the opportunity for interaction. At certain times, a few significant fluctuations were also seen among the trajectories. For PKM2-ligand complexes, these fluctuations were first observed before 0.02 ns follow by sharp increase to 0.155 with gradually decreased from to 0.155nm and then suddenly increased to 0.13 nm for F2-PKM2 Other than these we observed fluctuations at 0.08 ns and 0.1 ns. Similar results was observed for FDA standard although with lower magnitude. Whereas for F1-MCL1 fluctuations were first observed before 0.02 ns follow by sharp increase to 0.132 follow by decrease. Similar results was observed for FDA standard although with lower magnitude. The protein-ligand system likely experiences significant or minor conformational changes throughout dynamics simulation, as indicated by shifts in RMSD towards broader ranges. However, RMSD values that undergo lower ranges are ideally acceptable for protein systems [21]. The fluctuations observed for every residue signify the levels of flexibility obtained by them. Therefore, residues or a collection of residues with higher RMSF values suggest greater flexibility, which in turn suggests greater possibility for interaction with the ligand molecule. In the RMSF graph, the x-axis represented residues, while the y-axis represented fluctuation values. Residues obtained strong peaks, and considerable changes were seen in the area surrounding the active site. Additionally, many notable peaks of elevated fluctuations were also seen in the residues other than the active site residues, demonstrating their higher interaction potential and suggesting that the ligands could adapt effectively in the protein's binding pocket. The protein-ligand system likely experiences significant or minor conformational changes throughout dynamics simulation, as indicated by shifts in RMSD towards broader ranges. However, RMSD values that undergo lower ranges are ideally acceptable for protein systems [22].

The more negative coulomb energy between these interactions indicated good attraction between them. Both FDA approved drug had a more negative charge, compared to phytochemicals, and must have a less negative electrostatic energy. Our results confirms the subject. More hydrophobic contacts were indicated by LJ (short-range) energy values that were more negative. A solvent accessible surface area (SASA) plot was constructed from each system’s trajectory in order to analyze the surface area that is accessible to solvent for each system. their trajectory. The SASA information will be helpful to analyze, whether the ligand retained inside the shallow binding pocket or it expels out from the binding cavity. According to the result obtained, at the beginning of the simulation before 0.01ns all the interacting ligands retained inside the shallow binding pocket indicated by low magnitude of SASA till 100ps. This suggests that these molecules are less exposed to solvent. The decrease at the beginning shows that they have the properties of forming a stable complex with this main protease enzyme.

it was observed that Mormordicoside F1 and F2 follow Lipinski’s rule towards drug-likeness with two H-bond donor (not more than 5), five H-bond acceptor (not more than 10), A logP value of 4.317(not more than 5), rotatable bonds less than 10), Topological Polar Surface Area (TPSA) off 117.84 Å2 (< 140 Å2 ), and The low logP value of both ligand indicated good absorption and permeation with higher hydrophilicity. Additionally, it was found to be non-cancerous, HIA positive, BBB positive, non-AMES toxic and Caco-2 permeable as well as having negative hERG inhibition, negative results for aromatase binding, micronuclear, and negative results for biodegradation. HIA value of F1 and F2 phytochemicals was found to be than the FDA approved standard (Venetoclax and Benserazide). Zero LD50_oral, Genotoxic_ Carcinogenicity_Mutagenicity and SureChEMBL value were recorded for both F1 and F2. The two main Cytochrome P450 enzymes that are important for drug metabolism in the liver are CYP2D6 and CYP3A4. The analysis identified F1 and F2 as a CYP2D6 substrate/ non-inhibitor and CYP3A4 non-substrate/non-inhibitor, indicating that the drug may be metabolized in the liver. Mormordicoside F1 and F2 was also identified as a P-glycoprotein nonsubstrate/non-inhibitor; therefore, it may not be easily transported in the body. The quantitative estimate drug-likeness value of 0.283 The compound was identified with few toxicities such as acute aquatic toxicity, Skin_Sensitization etc., but the values were found to be mostly lower than a few of the clinical drug .

The phylogenetic relationship between the protein target sequences and the corresponding selected homologs for PKM2, infers that a likely comparable result from the docking bond affinity can be generated with 5X1W, 5X1V and 4G1N based on the high bootstrap confidence and might not be similar for 6V74, 6V75 and 6V76 protein target due to the relatively low bootstrap confidence. MCL1 protein target sequences phylogenetic relationship shows similarity to all the protein sequence implies likelihood of reoccurring result of the Insilco profiling.

With their role in tumor development, PKM2 and MCL1 acted as a promising target, whose inhibition could potentially control tumorigenesis. To this end, we attempted to investigate the binding patterns of the bioactive molecules of Mormordica charantia by docking them against the active site of PKM2 and MCL1. This exploration uncovered the bioactive molecules Mormordicoside F2 interaction with PKM2 to have achieved the best ligand-receptor structure from the docked based on the lowest energy and reasonable number of Hbonds formed between the target and ligand. From this study, Mormordicoside F2 could be considered as efficient ligand targeting Pyruvate kinase muscle isozyme 2. This binding of Mormordicoside F2 to PKM2 could facilitate formation of tetramer,Constitutively activating PKM2 and tumorigenesis suppression, thereby preventing cancer initiation and development. Similarly, F1-MCL1 interaction exhibited the best ligand-receptor structure among MCL1 complexes from the docked based on the lowest energy and reasonable number of Hbonds formed between the target and ligand. Mormordicosid F1 shows potential to regulated mitochondrial apoptosis controlled by members of B-cell leukemia/ lymphoma-2 (BCL2) family proteins via their homotypic and heterotypic interactions, thereby preventing cell proliferation to outpace apoptosis. The admet analysis results obtained identified the potential of Mormordicoside F1, F2, Ben.serazide and Venetoclax as a promising drug candidate after the comparison of all the ADMET properties of this phytochemicals with the FDA standard drugs of the study.

Furthermore, molecular dynamics simulation was conducted to gain a molecular understanding of the self-conformational perturbations that the protein causes in order to achieve conformational flexibility and stability with the ligand. After the MDS and refinement, RMSDs obtained shows that all the samples were of very low magnitude, this results are a strong indication of conformational stability that the protein had achieved with the ligand molecules. The overall energy of all the complexes complex has also been stabilized with all the structures having a good root-mean-square deviation (RMSD) score. The more negative coulomb energy (short-range) and LJ (short-range) observed between these interactions indicated good attraction between them. The results obtained from SASA analysis shows that Mormordicose F1, F2, Venetoclax and Ben.serazide ligands retained inside their shallow binding pocket. However, Molecular dynamics and simulation at higher steps and simulation time need to be performed for more vital information. The study also needs in-vitro with in-vivo animal studies for the confirmation of these phytochemicals as a potent inhibitor of MCL1 and PKM2 towards cancer treatment.

MCL1: Anti-apoptotic Myeloid leukemia 1

PKM2: Pyruvate kinase muscle isozyme 2

SASA: solvent-accessible surface area

RMSD: Root-mean-square deviation

RMSF: Root-mean-square fluctuation

STAT3: Signal transducer activator of tra

Nscription 3

AKT2: RAC-beta serine/threonine-protein kinase

B-NHL: B-cell non–Hodgkin lymphomas

PEP: Phosphoenolpyruvate

Acknowledgements

All authors are highly acknowledged for their contributions.

Author contributions

TAI & JOA: conceptualization, data curation, formal analysis, investigation, supervision, validation, methodology, project administration, software, visualization, writing – original draft. TIA, JOA, PIO, DA, AGA & FIA: investigation, project administration, software, supervision, validation, writing – review & editing. All authors read and approved the final manuscript.

Funding

Not applicable.

Availability of data and materials

All data and materials are available upon request.

Ethical approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

There is no conflict of interest regarding publication of the current research work.

Ferlay J, Ervik M, Lam F, Colombet M, Mery L, Piñeros M, et al. Global Cancer Observatory: Cancer Today. Lyon: International Agency for Research on Cancer; 2020 (https://gco.iarc.fr/today, accessed september 2021).
Dayton TL, acks T, Vander M J. PKM2, cancer metabolism, and the road ahead EMBO reports 2016; 17:12 https://doi.org/10.15252/embr.201643300
Wang, F, Yang Y. Inhibition of PKM2 sensitizes triple-negative breast cancer cells to doxorubicin Biochem Biophys Res Commun 2014; 454:465–470. https://doi.org/10.1016/J.BBRC.2014.10.109
Liang J, Cao R, Wang X, Zhang Y, Wang P, Gao H, Li C, Yang F, Zeng R, Wei P, et al. Mitochondrial PKM2 regulates oxidative stress-induced apoptosis by stabilizing Bcl2 Cell Res 2017;27:329–351. https://doi.org/10.1038/cr.2016.159
Amin, S, Yang, P, and Li, Z. Pyruvate kinase M2: a multifarious enzyme in non-canonical localization to promote cancer progression Biochim Biophys Acta Rev Cancer 2019; 1871:331–341. DOI: 10.1016/j.bbcan.2019.02.003
Chen XS, Dongsheng YU Protein kinase function of pyruvate kinase M2 and cancer Cancer Cell Int 2020; 20:523 https://doi.org/10.1186/s12935-020-01612-1
David, CJ, Chen, M, Assanah, M, Canoll, P, and Manley, JL.HnRNP proteins controlled by c-Myc deregulate pyruvate kinase mRNA splicing in cancer Nature 2010; 463, 364–368. https://doi.org/10.1038/nature08697
Zahra K, Dey T, Ashish, Mishra SP and Pandey U. Pyruvate Kinase M2 and Cancer: The Role of PKM2 in Promoting Tumorigenesis Front Oncol 2020; 10:159 doi: 10.3389/fonc.2020.00159
Puckett, D.L, Alquraishi, M, Chowanadisai, W, Bettaieb, A The Role of PKM2 in Metabolic Reprogramming: Insights into the Regulatory Roles of Non-Coding RNAs Int J Mol Sci 2021; 22, 1171 https://doi.org/10.3390/ijms22031171
Hsu MC, Hung WC Pyruvate kinase M2 fuels multiple aspects of cancer cells: from cellular metabolism, transcriptional regulation to extracellular signaling Mol Cancer 2018; 17:35 https://doi.org/10.1186/s12943-018-0791-3
He Y, Gao M, Cao Y, Tang H, Liu S, Tao Y Nuclear localization of metabolic enzymes in immunity and metastasis Biochim Biophys Acta 2017;1868:359–71. http://dx.doi.org/10.1016/j.bbcan.2017.07.002
Wang H, Guo M, Wei H, Chen Y. Targeting MCL-1 in cancer: current status and perspectives. Wang et al J Hematol Oncol 2021; 14:67 https://doi.org/10.1186/s13045-021-01079-1
Klanova M, Dmitry K, Pokorna E, Zikmund E, Karolova J, Behounek M, Renesovo N, Sovilj D, Cristina DK, Helman K, Radek J, Ondrej H, Ladislav A, Marek T, Pavel K. Anti-apoptotic MCL1 Protein Represents Critical Survival Molecule for Most Burkitt Lymphomas and BCL2-negative Diffuse Large B-cell Lymphomas. Mol Cancer Ther 2022; 21:89–99 doi: 10.1158/1535-7163.MCT-21-0511
Al-Odat OS, von Suskil M, Chitren RJ, Elbezanti WO, Srivastava SK, Budak-Alpddogan T, Jonnalagadda SC, Aggarwal BB and Pandey M (2021) Mcl-1 Inhibition: Managing Malignancy in Multiple Myeloma Front Pharmacol 2021; 12:699629 doi: 10.3389/fphar.2021.699629
Hird AW, Tron AE Recent advances in the development of Mcl-1 inhibitors for cancer therapy Pharmacol Ther 2019; 198(06):59–67 https://doi.org/10.1016/j.pharmthera.2019.02.007
Shahar N, Larisch S Inhibiting the inhibitors: targeting anti-apoptotic proteins in cancer and therapy resistance Drug Resist Update 2020; 52(09):100712 https://doi.org/10.1016/j.drup.2020.100712
Ngbolua K, Kilembe JT, Matondo A, Ashande C, Mukiza J, Nzanzu CM, Ruphin F, Robijaona B, Mpiana PT, Virima M. In silico studies on the interaction of four cytotoxic compounds with angiogenesis target protein HIF-1α and human androgen receptor and their ADMET properties JeanPaul KotoTeNyiwa. Bulletin of the National Research Centre 2022; 46:101 https://doi.org/10.1186/s42269-022-00793-1.
Selvaraj A, Antony S, Hakdong S. Anti-methanogenic efect of rhubarb (Rheum spp.)-An in silico docking studies on methyl-coenzyme M reductase (MCR). Saudi J Biol Sci 2019; 26:1458–1462. https://doi.org/10.1016/j. sjbs.2019.06.008
Vidya M, Madhu SS, Batoul F, Mastan M, Afroz A, Ganj PN (2019) Molecular docking of angiogenesis target protein HIF-1α and genistein in breast cancer. Genetics 701:169–172. https://doi.org/10.1016/j.gene.2019.03.062
Olowosoke CB, Alaba AA, Adegboyega BB. Citrullus lanatus natural product library: A hoard of viable potential inhibitor candidates for diabetes mellitus type II therapeutic target enzymes World Journal of Advanced Research and Reviews, 2022, 15(01), 534–560 https://doi.org/10.30574/wjarr.2022.15.1.0713
Sharma J, Bhardwaj VK, Rahul S, Vidya R, Rituraj P, Sanjay K. An in-silico evaluation of different bioactive molecules of tea for their inhibition potency against non-structural protein-15 of SARS-CoV-2 Food Chemistry 2021; 346:128933 https://doi.org/10.1016%2Fj.foodchem.2020.128933
Chatterjee S, Maity A, Chowdhury S, Islam MD, Muttinini RK, Debanjan S. In silico analysis and identification of promising hits against 2019 novel coronavirus 3C-like main protease enzyme, Journal of Biomolecular Structure and Dynamics, 2020; DOI: 10.1080/07391102.2020.1787228.

Table 1 is available in the Supplementary Files section.

Download PDF

Version 1

posted

You are reading this latest preprint version

Classes of Mormordicoside with potent and selective tumor cell growth inhibitory activity: prediction of Pyruvate kinase muscle isozyme 2 (PKM 2) and Anti-apoptotic Myeloid leukemia 1 (MCL1) inhibitor through machine learning.

Status:

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

Background

Methods

Results

Discussion

Conclusion

Abbreviations

Declarations

References

Table 1

Supplementary Files

Status:

Version 1